Introduction

Rivers are the most important life supporting systems of nature. For centuries, humans have been enjoying the natural benefits provided by rivers without understanding much on how the river ecosystem functions and maintains its vitality (Naiman 1992; Naiman and Bilby 1998). Man has changed the nature of many of the world’s rivers by controlling their floods, constructing large impoundments (Ittekkot and Lanne 1991), overexploitation of living and non-living resources (Macfarlane and Mitchell 2003; Kitetu and Rowan 1997) and using rivers for disposal of wastes (Haslam 1990). Among these, indiscriminate extraction of non-living resources like sand and gravel from riverbed is the most disastrous as this activity threatens the very existence of the river ecosystem (Kondolf 1994; Poulin et al. 1994). A review of literature reveals that indiscriminate extraction of river sand and gravel many folds higher than natural replenishments can impart serious offsite and onsite impacts. This ultimately leads to changes in channel form, physical habitats and food webs (Starnes 1983; Rivier and Seguier 1985; Thomas 1985; Sandecki 1989; Kitetu and Rowan 1997), engineering structures associated with river channels (Bull and Scott 1974; Harvey and Smith 1998) and inland sediment supply to coastal and nearshore environments (Jenkins et al. 1988; Gaillot and Piegay 1999). As these adverse effects become increasingly recognised and understood, instream sand mining/aggregate extraction has recognised increasing scientific scrutiny (Bates 1992; Collins 1995; Norman et al. 1998; Kondolf et al. 2002). Although more focussed researches leading to restoration of river environments are progressing in many developed countries (Tompkins and Kondolf 2003; Kondolf 2006; Kondolf et al. 2006), much attention has not been made in the rest of the world.

Here we report the environmental effects of river sand mining from the small rivers (catchment area < 10,000 km2; Milliman and Syvitski 1992) debouching into the largest marginal lagoon, in the southwest coast of India, the Vembanad lake. These rivers are being subjected to indiscriminate sand extraction over the past 3–4 decades consequent to economic development, rise in foreign remittance and liberalised housing schemes from banking sector for building constructions since early 1970s. Studies (Padmalal and Arun 1998; Padmalal et al. 2006) reveal that sand mining is taking place many folds higher than natural replenishments, which, in turn, led to severe environmental degradation and ecological disorders. Reduction in sediment supply from upstream reaches and erosion of its own channel during high flow regimes of monsoon season are common in many rivers in the southwest coast. This has led to channel incision and undermining of engineering structures like bridges, side protection walls, water intake structures constructed for rural water supplies, etc. Lack of sufficient information regarding the negative effects of sand mining is a major set back for laying strategies for the conservation and management of the small catchment rivers of the region. Considering the importance of scientific assessments on the extent of environmental degradation consequent to indiscriminate sand mining, an attempt has been made in this paper to address a few aspects about the impacts of sand mining on the physical and biological environments of Kerala rivers in the southwest coast of India, taking the river catchments of Vembanad lake as an example.

Geology and environmental setting

The Vembanad lake, the largest coastal lagoon (∼250 km2) in the west coast of India, and its catchments cover an area of about 14,500 km2. The entire area lies between North latitudes 9°0′–10°35′ and East longitudes 76°05′–77°25′ (Fig. 1). The lake catchment is drained by seven small rivers (i.e. rivers with catchment area <10,000 km2; Milliman and Syvitski 1992) viz., Chalakudy, Periyar, Muvattupuzha, Meenachil, Manimala, Pamba and Achankovil rivers. These rivers originate from the Western Ghat mountains and flow generally westwards before debouching into the Vembanad lake.

Fig. 1
figure 1

Study area showing Vembanad lake and river catchments

Full size image

Physiographically, the study area can be divided into 3 distinct zones—the highlands (>75 m above msl), midlands (8–75 m) and the lowlands (<8 m); (CESS 1984). The total population of the area as per 2001 census is about 10 million and population density about 690 inh km−2. The major rock types in the area are: (1) Precambrian crystallines, (2) Tertiary sedimentaries and (3) Recent to Sub-Recent sediments. The Precambrian crystallines cover about 85% of the study area and composed mainly of charnockites, khondalites, garnet-biotite gneisses and hornblende gneisses (GSI 1995). The Tertiary sedimentaries are represented by Quilon and Warkalli Formations. While the former (i.e. the Quilon Formation) is composed of limestones, sandstones and claystones, the latter is formed of sandstones and claystones with intercalations of lignite seams. The Precambrian crystallines and the Tertiary sedimentaries are lateritised at many places in the midlands. Recent to Sub-Recent sediments of Late Quaternary age, represented by coastal sands and alluvium, are confined to the lowlands and parts of midlands (Padmalal et al. 1997). The area experiences a tropical humid climate with an average annual rainfall of about 3,300 mm. Of the total rainfall, southwest monsoon (June–September) contributes about 75% and the remaining by northeast monsoon (October–December) and summer showers. The temperature varies from 20 to 39°C. Humidity is high (>80%) during monsoon period.

Rivers in the Vembanad lake catchments

Table 1 shows the characteristic features of the Achankovil, Pamba, Manimala, Meenachil, Muvattupuzha, Periyar and Chalakudy rivers draining the Vembanad lake catchments. The headwater elevation of these rivers varies between 700 and 1,830 m with respect to mean sea level. These small, mountainous rivers (Milliman and Syvitski 1992), except the Achankovil, which is an upland river, enfold alluvial deposits rich in construction grade sand and gravel in its lowlands and midlands. The highlands, which act as the main production zone for sediments, are generally under dense forests and/or forest plantations. A substantial proportion of the sediments derived from the uplands is being trapped in the reservoirs constructed for irrigation and hydroelectric power generation. The channel width of the mainstream that holds thick sand beds in the active channel ranges from 100 to 500 m. These rivers generally exhibit higher meandering index in the midlands than the other two physiographic zones. The alluvial reaches exhibit a variety of roughness configurations that vary with slope and position within the channel network. The entire alluvial reaches of these rivers can be categorised into three distinct morphological classes towards downstream—plane bed reach, pool riffle reach and dune ripple reach. The content of gravel and sand exhibits marked changes over these morphological classes—the former (i.e. gravel) decreases in percent contents, while the latter exhibits a marked increase downstream.

Table 1 Salient features of the rivers draining the Vembanad lake catchments
Full size table

Methodology

The study includes extensive field survey in the entire channel networks of the Vembanad lake catchments for: (1) mapping locations of river sand mining and storing centres, reaches of river degradation, locations of damaged engineering structures like side walls, bridges, spillways and water intake structures, etc., and (2) collection of primary and secondary data on mining operations, quantity of sand and gravel extracted from each locations and information on biological resources of the respective rivers. The current extraction of sand from the river stretches falling within each river basin was estimated from number counts and select weighing of the loaded vehicles moving out the mining locations during the period of sand extraction, interviews with local people and labourers, secondary information available with the district and local administration etc using standard formats (Gilpin 1995). The information on the discharge of sand and riverbed lowering was collected from the river gauging stations of the Central Water Commission (CWC), Government of India. The data collected through the major ongoing research programmes of Centre for Earth Science Studies (CESS) and Centre for Water Resource Development and Management (CWRDM), Department of Mining and Geology, Government of Kerala, Non Governmental Organisations (NGOs) like Pampa Parirakhshana Samithi (PPS), All Kerala River Protection Council (AKRPC), etc., were the other sources of information for the present study. The data analysis and interpretations were done following the published works of Kondolf et al. (2002), Kitetu and Rowan (1997) and Weeks et al. (2003).

River sand mining

River sediments comprise a spectrum of particle sizes such as boulders, cobbles, pebbles, granules, sand, silt and clay. In general, coarser particles concentrate in the low order, high gradient streams and finer particles progressively deposit towards the downstream because of reduction in shear stress and improved sediment sorting (Kondolf 1994). Three sediment delivery processes are recognised, (1) mass wasting on hill slope, (2) hill slope erosion and (3) erosion of channel bed and banks (Collins and Dunne 1990). Mass wasting includes landslides and soil creep, and occurs when gravity alone moves soil to the river channels. Hill slope erosion occurs when precipitation intensity exceeds the absorption capacity of soil and generates overland flow. In relatively undisturbed river systems at its mature phase, gradual erosion of outside bends of meanders and deposition of eroded materials on inside bends cause an imperceptible shifting of channel within its floodplain. This form of stability is called dynamic equilibrium (Heede 1986). Although river flows and sediment loads are variable within and among years, sediment balance and channel stability occur over long term. Instabilities introduced by human activities like deforestation, sand and gravel mining and other activities, and by natural processes like extreme precipitation, forest fires and other events can cause channel bed and banks to become net sources of sediment.

River sand and gravel are mined extensively from the drainage networks of the Vembanad lake catchments, irrespective of the physiographic zones, river orders and ecological significance. However, the intensity of mining is high in the alluvial reaches of the main channels. In addition to mining of sand and gravel from active channels (instream mining), a substantial amount of sand is also being extracted from the overbank areas (floodplain mining) of the river as well. Generally, two types of instream mining are in practice in the area—pit excavation and bar skimming. Pit excavation is extraction of sand and gravel from the riverbed or floodplain areas by uncontrolled digging. In some cases, diesel powered suction pumps are used to extract sand rich sediments from wet pits in the active channels and floodplains. Bar skimming, on the other hand, is the controlled extraction of sand from the exposed sand bars (instream bars and point bars) in the channel environment. Usually, bar skimming would be done above the water table and within a minimum width buffer that separates the excavation site from the low flow channel and the adjacent active channel bank. Of the two types, pit excavation is the widespread sand mining method adopted in the alluvial reaches of the rivers in the Vembanad lake catchments.

Table 2 shows the estimated quantity of sand extraction from various rivers of the study area. It is revealed that an amount of 11.73 million tonnes of sand and gravel are being extracted annually from the active channels of these rivers. Out of the total quantity of sand extraction, nearly half the amount (i.e. 5.64 million tonnes) is from the Periyar river, which flows through the most urbanized and industrialized areas of the Vembanad lake basin, the Kochi city. Although small in river length and catchment area compared to Periyar river (244 km/5,398 km2), the river Muvattupuzha (121 km/1,554 km2) is also affected severely by sand extraction (2.00 million ty−1), owing to its proximity to the development centre, the Kochi city and its satellite townships. The sand extraction in the other rivers are of the order Manimala > Achankovil > Chalakudy > Pamba > Meenachil (see Table 2). Spatial analysis reveals that about 75% of the extracted sand is from the midlands, which cover only 30% of the entire study area. The highlands and lowlands account for almost equal amount of sand extraction (Table 2), although the former covers a greater portion of the study area (60%) than the latter (10%). The sand harvesting (mining) centres are also higher in numbers in midlands (523) than highlands (309) and lowlands (171). A total of about 22,300 registered labourers are employed in instream mining sector and the number of indirect workers depending on the quarried sand, especially in the construction sector, would be over a lakh. In some places, sand mining is even clubbed with extraction of subfossil logs embedded within alluvial sand deposits (Fig. 2). These subfossil logs that are segregated in the Early Holocene higher sea levels (Kumaran and Nair 2005) are used as firewood in the brick kilns of the nearby areas. An apparent assessment of the quantity of sand extraction with respect to the available sand input data reveals that sand mining in the river reaches down to the gauging stations is of the order of 40 times higher than the sand input recorded at the gauging stations of the respective rivers (Table 3). The natural replenishment is very meagre as the river catchments are occupied mainly by Precambrian crystallines. Further, a substantial part of sand originated from natural crustal weathering will be detained within the reservoirs constructed in the uplands for irrigation and hydropower generation (see Fig. 1). According to Nair and Padmalal (2006) whatever sand reaches in the alluvial zones are derived from erosion/slumping of early–middle Holocene sand layers interbedded within the river banks of the midlands and lowlands.

Table 2 Sand mining and other relevant details of the rivers draining the Vembanad lake catchments with respect to different physiographic zones
Full size table
Fig. 2
figure 2

Combined mining of subfossil logs and river sand: a scene from the Muvattupuzha river

Full size image
Table 3 Quantity of river sand mining with respect to the sand discharge estimate of the gauging station
Full size table

River degradation, damage to agricultural lands close to river channel, etc., led to agitation among the local people who directly depend on rivers for various purposes. People’s opposition and strict regulation of instream sand mining often lead to mining of sand from floodplains and/or river terraces. Floodplain mining is severe in the lowlands of Muvattupuzha river basin compared to the others. It is estimated that an amount of 0.414 million ty−1 of sand and gravel is being mined annually from the floodplain areas of the seven rivers subjected to the present study (Table 4). The number of pits and labour force in the floodplain-mining sector are 137 and 600, respectively.

Table 4 Estimated quantity of floodplain mining in the rivers joining the Vembanad lake
Full size table

Effects of instream sand mining

River basins have long been exploited as sources of fine aggregates for building constructions. Depending on the geologic and geomorphic setting, river sand mining can impose serious environmental consequences in the long run (UNEP 1990). The cumulative effects of uncontrolled sand mining have substantially altered the physical, chemical and biological environments of the small southwest Indian rivers. Table 5 shows some of the general impacts of river sand mining noticed in the river catchments of Vembanad lake. Degradation of rivers is severe in the alluvial reaches of the midlands and lowlands compared to the production zones in the highlands. In some cases, the riverbank itself is being scooped out first for brick earths and then the intervening sand layers for fine aggregates.

Table 5 Impact of river sand and gravel mining on various environmental components of rivers draining the Vembanad lake
Full size table

It is now well understood that indiscriminate sand mining well above the natural replenishments could lead to serious environmental problems in the river basins. Of the two types of sand mining from the active channels described earlier, pit excavation is the most dangerous than bar skimming. Excavation of mining pits in the active channel of the alluvial reaches lowers the streambed, creating knick points. The knick point could locally steepen the channel slope, which in turn increases the flow energy. During high flows of the monsoon season, knick point becomes a location of bed erosion and gradually moves upstream through head cutting (Garde and Raju 1985; Collins and Dunne 1989; Kondolf 1994). This process could mobilize a substantial quantity of streambed sediments that are then transported downstream and ultimately deposited in the excavated area. In some cases, the headcuts moves long distances upstream and often spreads into even the tributary channels (Hartfield 1993; Kondolf 1997). Merging of the adjacent pits during monsoon season could lower the streambed faster imposing instability to the riverbanks and riparian vegetation. The sediment-laden water passing downstream leaves coarser sediments like gravel and sand in these pits and thereby carries only sediments much less than its capacity. Thus, the hungry water leaving the pit may erode the downstream reaches with aggression, ultimately ending up in the degradation of the entire river channels. A similar kind of process in the Cheny creek in USA, degraded the river stretch, ultimately lowering the stream bed as much as 4.9 m from the original bed level (Lane 1955). It is reported that pit excavation in the active channel of a Russian river near Healdsburg, California in 1950s and 1960s caused channel incision in excess of 3 m over a 11 km length of river, with a maximum incision of 6 m (Collins and Dunne 1990; Kondolf 1997). A similar effect occurs below dams, which trap sediment and release hungry water downstream where channel incision usually ensues; instream sand excavation below dams compounds this problem (Kondolf and Swanson 1993; Kondolf and Larson 1995).

All the seven rivers draining the Vembanad lake catchments are degrading severely since 1980 consequent to indiscriminate mining of construction grade sand from instream and floodplain areas. As a result, the channel beds in the storage zones of these rivers are lowering fast posing instability problems to riverbanks and associated engineering structures. Table 6 shows the rate of channel bed lowering recorded by the gauging stations of the Central Water Commission located in the storage zones of these rivers. Among the seven rivers, the Periyar river draining the most industrialised and urbanised area of Kerala state, the Kochi city, records the highest bed lowering value of 19 cm year−1. Furthermore, the rate of bed lowering exhibits a marked increase in the last 2–3 decades (Table 6; Fig. 3). Contrary to Periyar, the Achankovil river draining the southern part of the study area exhibits a reverse trend with a marked decrease in bed lowering during 1985–2000 period. This is because, the channel bed at the gauging station of Achankovil river can no longer respond to sand mining as the channel here has already touched the crystalline basement rocks. The rate of channel bed lowering in the other rivers exhibits wide fluctuation during the monitoring periods because of interventions of Honourable Court and Government from time to time. Although such interventions could yield positive results for short-term periods (which in turn reflected well in the slow pace of channel bed lowering at certain periods) the efforts could not sustain long due to the ever increasing demands of river sand for construction works. The riverbed lowering and channel degradation have damaged many bridges, rural water supply schemes and side protection structures in addition to imposing habitat changes through riparian and instream vegetation (Fig. 4). Many perennial wells in the floodplain areas became dried up in summer season due to lowering of water table in the nearby areas of the river channel.

Table 6 Riverbed lowering (cm year− 1) consequent to indiscriminate extraction of sand from the rivers draining the Vembanad lake
Full size table
Fig. 3
figure 3

Riverbed lowering of Periyar river during the period 1980–2000 (Data source: CWC, Kochi)

Full size image
Fig. 4
figure 4

A water intake structure constructed in Manimala river is found exposed due to river bed lowering, which is an after effect of sand mining

Full size image

Apart from the above changes in the physical environment and damages to engineering structures, indiscriminate sand mining could also impose drastic changes in the biodiversity of aquatic fauna and flora. According to Kurup et al. (2005), the physical habitat variables play a leading role in the distribution of many fish species of Pamba, Periyar and Chalakudy rivers. The study revealed that the habitat alteration consequent to indiscriminate sand mining contribute significantly to the endangerment of fishes like Labeo dussumieri, Puntius denisonii, Batasio travancoria, Horabagrus nigricollaris and Lepidopygopsis typus. In another study in the Achankovil river, Sunilkumar (2002) reveals that the benthic fauna in the river has severely affected by indiscriminate river sand mining. Number counts of benthic animals were very low or even absent in some cases, in the sampling stations affected severely by sand mining (Fig. 5). Scooping of sand along with the larval forms of many insects like mayfly, dragonfly, caddisfly and other insects of the order Diptera has imposed marked decrease in population of these beneficial insects which form food to other higher order animals like fishes and amphibians. Loss of riparian habitats due to caving of river banks (Fig. 6), river bed lowering and also clearing of vegetation for brick earth and sand mining for the development of roads (Fig. 7), sand harvesting centres, etc. adversely affected the ecosystem functioning of the river reaches in midlands and lowlands. Loss of instream vegetation, sand and gravel in the river bed, etc., not only affect the feeding and hiding grounds of fishes and other freshwater animals but also induce spawning disorders to phytophilic and psammophilic fishes in the aquatic environment (FAO 1998). This may, of course, end up in the decline of inland fishery resources, which ultimately affect the protein and energy flow to fish feeding animals in the aquatic and terrestrial environments.

Fig. 5
figure 5

Average abundance (0.1 m2 area) of benthic organisms in the Achankovil river (Data source: Sunilkumar 2002)

Full size image
Fig. 6
figure 6

Slumped/exposed bank of Pamba river due to unrestricted mining activities

Full size image
Fig. 7
figure 7

Unplanned roads for transportation of vehicles into the river bed for easy loading of the mined sand along the river channels of Kerala is a common scene during past several years

Full size image

Conclusions and recommendations

In the past 3–4 decades, rivers in the densely populated areas of the world are subjected to immense pressures due to various kinds of human interventions, among which indiscriminate mining for construction grade sand from alluvial reaches is the most disastrous one. This is mainly because of the fact that uncontrolled scooping of sand aggravates river degradation and threatens its trophic structure. The situation is rather alarming in the small rivers of Kerala in the southwest coast of India, which support the life and greenery of the region. Loss of riparian and instream vegetation, changes in the feeding, breeding and spawning grounds of aquatic organisms including fishes not only impose stress in the river ecology but also create damages in the terrestrial and nearshore marine environments as well. All these points to the immediate need for multidisciplinary studies in the riverine environments of the Vembanad lake for assessing the extent of degradation consequent to indiscriminate sand and gravel mining. This is very much important for laying down strategies for regulating the mining activities on environment friendly basis and also for creating awareness on the impact of river sand mining on the physical and biological environment of these life support systems.

Given below are some of the recommendations suggested to improve the overall environmental quality of the river systems of southwest coast of India, in general and Vembanad lake catchments, in particular.

  • An integrated environmental assessment, management and monitoring program should form part of the sand extraction processes. Also, there is an urgent need for integrating the studies on various disciplines on the human induced degradation of the small catchment rivers of southwest coast of India.

  • Evaluate physical, chemical and biological effects of instream mining on a river basin scale, so that cumulative effects of extraction on the aquatic and riparian resources can be recognised and addressed at various levels for proper remedial measures.

  • Examine and encourage alternatives to river sand for construction purposes. Immediate steps are to be taken to intensify research activities leading to the finding of a suitable, low cost and easily available alternative to river sand.

  • Evaluate control measures such as bank stabilisation, revegetation of buffer strips, influences of connected floodplain pits etc. Restoration efforts should concentrate on techniques that will optimise fish production, promote aquatic diversity and restore biotic integrity.

  • Awareness campaign should be conducted at various levels about river sand mining, present state of environment of rivers and immediate need for control measures.