Lacustrine sedimentation in a monsoon environment: the record from Phewa Tal, middle mountain region of Nepal

Lacustrine sedimentation in a monsoon environment: the record from Phewa Tal, middle mountain region of Nepal

Geomorphology 27 Ž1999. 307–323 Lacustrine sedimentation in a monsoon environment: the record from Phewa Tal, middle mountain region of Nepal Jamie R...

2MB Sizes 4 Downloads 36 Views

Geomorphology 27 Ž1999. 307–323

Lacustrine sedimentation in a monsoon environment: the record from Phewa Tal, middle mountain region of Nepal Jamie Ross, Robert Gilbert

)

Department of Geography, Queen’s UniÕersity, Kingston, Ontario K7L 3N6, Canada Received 1 January 1998; revised 23 July 1998; accepted 9 August 1998

Abstract A watershed-scale approach to the sedimentary environment of the Phewa Tal drainage basin is used to assess rates and patterns of sediment accumulation within Phewa Tal reservoir. The susceptibility of surficial materials to erosion and transportation is highest in the early part of the monsoon season but major erosional events also occur during the monsoon proper. Thus, sediment delivery to the reservoir is at least two orders of magnitude higher during the monsoon than the dry season. In the reservoir, suspended sediment is distributed mainly by interflows at 4 to 6 m and 8 to 12 m depth. Phewa Tal has four distinct sedimentary environments as interpreted from the sedimentary record in cores. Ž1. The deltaic environment, which receives sediment directly from the main inflowing stream, the Harpan Khola, and from up-valley wetlands and agricultural fields, experiences rates of sedimentation of up to 1 m ay1. Ž2. The delta-proximal region where clearly varved sediments are deposited in response to the annual monsoon, has a mean rate of sedimentation is 23.5 mm ay1. Ž3. The distal environment is beyond the influence of sediment-charged interflows, and is characterised by accumulation from suspension to form largely massive deposits, so that annual rates of sedimentation cannot be determined stratigraphically. Ž4. The region near the Pardi Dam is dominated by seasonal flushing and erosion of fine sediments during the monsoon. The sedimentary record documents the geomorphic processes of the drainage basin, especially as controlled by the monsoon, and indicates the impact of human occupance in the drainage basin in the twentieth century. The useable life of the reservoir is estimated as 360 years. q 1999 Elsevier Science B.V. All rights reserved. Keywords: erosion; lake sediment; varve; monsoon; Nepal

1. Introduction The physical properties and stratigraphy of lacustrine sediments are increasingly used as reliable proxies of climate, hydrology, and geomorphology in environments from tropical to polar. In particular, sedimentary records containing recognisable annual )

Corresponding author. Fax: q1-613-545-6122; E-mail: [email protected]

deposits are very useful because of the ease with which they can be measured and dated. Varved glacilacustrine sediments have been used to monitor long- and short-term changes in sediment delivery in glacierised catchments ŽPickrill and Irwin, 1983; Leonard, 1986; Gilbert et al., 1997. as influenced by factors on scales ranging from local, to regional, to global. Lacustrine sedimentation in a sub-tropical, monsoon environment, such as the Middle Mountains of

0169-555Xr99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 9 - 5 5 5 X Ž 9 8 . 0 0 0 7 9 - 8

308

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323

the Himalaya, is analogous to that in glacial settings. Rates of erosion and sediment delivery to lakes in both environments are extremely high. The input of water and sediment is strongly seasonal because of the occurrence of the monsoon rains in one environment, or to the summer melt of glaciers and snowpacks in the other. Thus, we anticipate that knowledge of glacial lakes has direct application to the monsoon environment, and that deposits in lakes and reservoirs of the Middle Mountains of Nepal can be interpreted to provide a record of the complex interactions between local climatic variability, natural erosional processes, and anthropogenic activity on a watershed scale. Recent research in the watersheds of Nepal, including the Jhikhu Khola ŽSchreier et al., 1994; Carver, 1997., the Likhu Khola ŽOverseas Development Agency, 1995., the upper Pokhara Valley ŽThapa and Weber, 1995. and the Phewa Tal basin ŽLohmann et al., 1988; Rowbotham, 1995. has documented the hydrologic and sedimentary systems of mountain catchments in this region and, in the case of Phewa Tal, the physical and biological limnology of the lakes within them. The early assessment that human activity is having a catastrophic effect on the environment has now been replaced by the contention of more limited human-related degradation ŽIves and Messerli, 1989.. With over 6000 rivers and a range in elevation from 60 to 8848 m asl, Nepal has more than 2.2% of the potential hydropower resources in the world and is second only to Brazil ŽShankar, 1991.. This potential has been partly realized through construction of the Kulekhani Reservoir near Kathmandu, and the Phewa Tal Reservoir at Pokhara. In response to concerns about rapid filling of these reservoirs by sediment, annual bathymetric surveys have been used by Leminen Ž1991., Sthapit and Leminen Ž1992., and Research and Soil Conservation Sections Ž1994a,b. to assess accumulation in these and other reservoirs in the region. However, these surveys do not relate sedimentation to processes in the drainage basin, nor to climatic influences or changes in land use. The purpose of this study is to assess lacustrine sediment as a record of the impact of natural and human-induced processes Žthe latter including dam construction, water diversion, agriculture, forestry

and urbanization. on erosion, transport and deposition of sediment in the Middle Mountains of Nepal ŽFig. 1.. This is accomplished through preliminary assessment of discharge and suspended sediment transport of the Harpan Khola Ž‘khola’ means ‘river’ in Nepali. as they vary between dry and monsoon seasons. This is followed by analysis of patterns of stratification and circulation in Phewa Tal Ž‘tal’ means ‘lake’ in Nepali. and discussion of sediment dispersal based on suspended sediment profiles collected after major storms. The sedimentary record of Phewa Tal is also analysed, and a model of sediment distribution patterns and rates of accumulation is proposed. The varved nature of the deposits in some parts of the reservoir is used to reconstruct the erosional and sedimentary environment of the Phewa Tal Watershed, and to determine the useable life of the reservoir. This watershed approach links the lacustrine environment with geomorphic and human processes in the drainage basin Žsee also Desloges and Gilbert, 1994, 1998. so that both may be better understood.

2. Methods Research was performed in the Phewa Tal Watershed during a portion of the dry season ŽSeptember–December. of 1995, and the monsoon season ŽJune–July. of 1996. A stream gauge was installed on the Harpan Khola at Pame ŽFig. 1. to monitor discharge, stage, and suspended sediment load during the monsoon. Stage was observed two or three times daily and converted to discharge using a rating curve based on five measurements ŽRoss, 1998.. In the Harpan Khola at Pame and in Phewa Tal, temperature and conductivity were measured with thermistor and conductivity probes in 1995, and these variables plus optical turbidity with a Hydrolab Datasonde 3 in 1996. Turbidity data from the Hydrolab ŽT in NTU. were calibrated with 0.5 l water samples filtered at 0.47 mm to determine suspended sediment concentration Ž c in mg ly1 . from the regression: c s 1.93 q 0.400 P T Ž r 2 s 0.969, n s 47. Žcf. Gilbert et al., 1997.. Cores of sediment were retrieved from the lake floor during both field seasons ŽFig. 1. using a modified percussion corer ŽGilbert and Glew, 1985.

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323

309

Fig. 1. Phewa Tal showing bathymetry and locations of coring sites and profiling stations. Insets show the drainage basin of the Harpan Khola and the location of the Middle Mountains Žshaded. in Nepal with the High Mountains and High Himalaya to the north, and the Siwaliks and the Terai to the south. Bathymetry modified from Leminen Ž1991..

with a 73-mm diameter tube, and a light-weight gravity corer having a 37-mm diameter tube. In the laboratory, magnetic susceptibility ŽMS. was measured with a Sapphire Instruments Model SI2B with a 95-mm external coil, and xradiographs were made. Cores were then split, allowed to dry, and photographed. All measurements below are expressed as lengths of wet core, although shrinkage and shortening of the cores occurred during drying. Detailed logs were recorded before the cores were sub-sampled for grain-size analysis Žcarried out by wet-sieving the sand fraction and analysis of the mud in a

SediGraph 5100.. Organic carbon and carbonate content were determined after Dean Ž1974. and Konrad et al. Ž1970. at 0.05- to 0.1-m intervals in selected cores by determining loss on ignition after combustion at 650 and 10008C, respectively. Thin sections were prepared following Lamoureux Ž1994..

3. Physical setting The Phewa Tal watershed Ž120 km2 . and reservoir Ž4.35 km2 . are located in the western part of the

310

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323

Pokhara Valley in the Middle Mountain physiographic region of central Nepal ŽFig. 1.. They are tributary to the Seti Khola which drains the southern slopes of the Annapurna Massif to the north. Elevations in the watershed range from 793 m asl, the normal operating level of the reservoir, to 2508 m asl at the western end of the catchment. Structurally weak, low-medium grade Precambrian to early Cambrian grey phyllitic schists underlie the basin in the north, with talc-rich, red phyllitic schists in the south ŽAnonymous, 1980.. The southwestern part of the basin contains quartzite schist interbedded with grey phyllitic schists, while the eastern end is dominated by carbonaceous conglomerate ŽImpat, 1980. derived from the Annapurna Massif ŽYamanaka et al., 1982.. Phewa Tal has a maximum depth of 22.5 m and a volume Žbased on recent bathymetric survey. of 37.76 = 10 6 m3 ŽResearch and Soil Conservation Sections, 1994a.. The lake is thought to have originally formed in the late Pleistocene, dammed by gravels transported from the Annapurna range by the Seti Khola ŽFort and Freytet, 1982; Yamanaka et al., 1982. which subsequently incised these deposits. A second natural dam formed between 1100 and 600 BP ŽYamanaka et al., 1982. from up to 5.5 km3 of material transported by a jokulhlaup in the Anna¨ purna range ŽCarson, 1985.. Conflicting reports exist on the dates of construction of artificial dams at Phewa Tal for hydroelectric generation and irrigation during the past 200 years, and of the subsequent failure ŽSharma, 1980; Ramsay, 1985; Science Applications International, 1993. but the following can be determined from these reports and from information obtained on site. In 1933 a stone and mortar dam was constructed ŽScience Applications International, 1993.. It was replaced in 1942 by a rockfill dam which failed in the 1960s. In 1967 the first concrete dam was constructed; it failed in 1975. The present dam became operational in June 1982 ŽRamsay, 1985.. Each dam failure exposed a considerable portion of the lake floor, although the exact history of water levels in the lake is unknown. The morphometry of the lake floor ŽFig. 1. reflects its history. The depression along the southern side is the former river channel, and the terrace to the north likely represents the floodplain of this river. Leminen Ž1991. divided the reservoir into three regions: the Silt Trap, the Main Reservoir, and

the River Channel. We further sub-divide the Silt Trap into Silt Traps I and II ŽFig. 1.. The climate of the Phewa Tal watershed is humid sub-tropical to humid temperate, with mean monthly temperatures from 12.8 to 25.78C. At Pokhara Airport Ž840 m asl. an average of 88% of the total annual precipitation Žmean, 3858 mm; maximum, 4843 mm in 1995. falls during the monsoon season of May through September. At Lumle Agricultural Centre Ž5 km northwest of the watershed, 1740 m asl. an average of 91% of the total annual precipitation Žmean, 5244 mm; maximum, 6219 mm in 1984. falls during the monsoon. The difference between the two stations reflects the positive relationship between precipitation and elevation ŽRamsay, 1985; Rowbotham, 1995., as well as the spatial variability in rainfall in the Middle Mountain region, and in the Himalaya in general. Almost 50% of the watershed is under agriculture. Crops include rice, maize, wheat, millet, barley, potato and mustard, with rice dominating lowland areas. What remains in forest is heavily harvested for fodder and fuel ŽRowbotham, 1995.. Part of the city of Pokhara occupies the rest of the basin along the eastern shore of Phewa Tal. Population that is growing at a rate of 1.9% per year ŽFleming, 1978. in urban Ž95,311 in Pokhara in 1991: Science Applications International, 1993. and rural Ž31,192 in 1991: Decore Consultancy Group, 1991. parts of the basin has placed increasing demands on the natural resources of the Phewa Tal watershed, and has altered its erosional and depositional environments.

4. Hydrology and limnology During June and July 1996, discharge in the Harpan Khola ranged from 11.1 to 58.6 m3 sy1 . Discharge was not measured during the dry season but data from Impat Ž1981. and Science Applications International Ž1993. suggest that mean flow rarely exceeds 10 m3 sy1 during this time. Measured suspended sediment values in the Harpan Khola varied from 0.7 to 3.8 mg ly1 during the dry season and from 8.3 to 157.8 mg ly1 during the monsoon season Žmeans, 2.1 and 50.4 mg ly1 , respectively.. The early portion ŽMay and June. of the monsoon season Žsometimes referred to as the pre-monsoon: e.g.,

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323

Carver, 1997. is the period of the greatest sediment yield. For example, discharge on 12 June was the lowest measured during the monsoon season Ž11.1 m3 sy1 . but the third highest concentration of suspended sediment Ž95 mg ly1 . occurred on this day. Heavy rainfall on dry, recently harvested fields led to a large fraction of total annual surficial erosion during this time Žcf. Carson, 1985; Carver, 1997.. Ross Ž1998. estimated sediment yield at the mouth of the Harpan Khola in Phewa Tal from 21 to 547 t dayy1 during the monsoon season compared to about 3 t dayy1 during the dry season, a range in mean loads of two orders of magnitude. Similarly, Carver and Nakarmi Ž1995. noted up to 90% of annual soil loss during the two largest events of the early part of the monsoon on erosion plots in the Jhikhu Khola watershed. Profiles of temperature and conductivity ŽFigs. 2 and 3. show that Phewa Tal was thermally stratified throughout the year until turnover in early December Žsee also Hickel, 1973; Rana, 1990.. However, during the monsoon on 19 July ŽFig. 2b., heavy rains

311

and elevated inflow partially broke down stratification, as also reported by Rana Ž1990.. Profiles of conductivity during the monsoon ŽFig. 3. mirrored temperature, with minima in the epilimnion, and maxima in the hypolimnion. As the monsoon progressed, conductivity in the epilimnion decreased as a result of dilution by monsoon rains and uptake of calcium by phytoplankton. Concurrently, conductivity increased in the hypolimnion as calcium was released by decaying phytoplankton ŽRoss, 1998.. Conductivity increased with distance from the river mouth, probably as a result of dissolution of carbonaceous conglomerates, the input of contaminated runoff from the city of Pokhara, and the diversion of water through the Seti Canal from the relatively carbonate-rich watershed of the Seti Khola, all occurring in the eastern part of the reservoir ŽFig. 1.. Profiles of suspended sediment trace particulate matter through the reservoir after major rainfall events. For example, profiles from three stations ŽFig. 4a., during a period of elevated input of sus-

Fig. 2. Temperature profiles from Ža. Station 3 ŽFig. 1. illustrating loss of thermal stratification from October to December, 1995 and Žb. Station 4 for June and July, 1996. Partial breakdown of thermal stratification occurred because of heavy rains on 19 July.

312

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323

Fig. 3. Conductivity profiles at Station 3 ŽJune and July, 1996. illustrating changes through the water column with the progression of the monsoon.

pended sediment on 13 June after a storm in the upper watershed, show sediment distributed down the former river channel ŽStations 3 and 4. and over the former floodplain ŽStation 10.. Concentrations of suspended sediment above 6 m were slightly higher at Station 3, because of the proximity of the river mouth and the effect of Coriolis force on the distribution of fine-grained particulate matter at these depths. A similar pattern at Station 4 indicates that interflows maintain structure up to 2 km from the river mouth, although lower concentrations show that significant settling from suspension had occurred. Heavy rains Ž109 mm at Pokhara Airport. on 17 and 18 July associated with the breakdown of thermal stratification discussed above resulted in the transport of significant amounts of debris to the reservoir. Turbidity values in excess of 1000 NTU Žthe upper limit of the sensor; over 400 mg ly1 . were recorded at Stations 1 and 2. All of the sediment entering Silt Trap I was derived from nearby wet-

lands and submerged agricultural fields. Much of the material delivered to the reservoir by the Harpan Khola entered Silt Trap II where it was subsequently deposited. Plumes of turbid water could also be seen along the southern shore of the lake. Concentrations of suspended sediment were sufficient to generate turbidity currents on the floor of the reservoir on 18 July ŽFig. 4b.. Initially, formation was observed at Station 11 where the concentration at the lake floor exceeded 200 mg ly1 . By the time flow reached Station 10, concentrations near the lake floor decreased to 50 to 60 mg ly1 These data also indicate that sediment derived from storms and associated mass movements in the watershed is distributed through much of the reservoir, and is not isolated to the former river channel. Although soils are most susceptible to erosion during the early part of the monsoon, major storms throughout the monsoon contribute a significant amount of the annual sediment yield to the reservoir. Much of the sus-

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323

313

Fig. 4. Profiles of the concentrations of suspended sediment showing Ža. interflows on 13 July, 1996, and Žb. underflows at Stations 10 and 11 on 18 July, 1996.

pended load is deposited in delta-proximal regions of the lake, as indicated by lower suspended sediment concentrations at more distal stations. 5. Sedimentary record Seven percussion cores and nine gravity cores from Phewa Tal ŽFig. 1. illustrate spatial variations in sediment distribution and accumulation. Sediments consist either of well-laminated silts and silty clays with a few sandy layers, or massive deposits that only change in colour from lighter to darker with depth. Based on detailed analyses of these deposits and the patterns of sediment distribution in the lake, Phewa Tal was divided into four sedimentary environments ŽFig. 5. described below. 5.1. Deltaic enÕironment The deltaic environment receives most of the input from upstream paddy fields and the Harpan

Khola. Bathymetric surveys by Sthapit and Leminen Ž1992. revealed rates of sediment accumulation in this area of up to 1 m ay1 , and an annual sediment load to both Silt Traps of 105,364 m3 during 1990– 1991. In 1990, the Harpan Khola entered the reservoir through Silt Trap I but it has since reoccupied the finger delta that is rapidly prograding to isolate Silt Traps I and II from the Main Reservoir. Comparison of its position mapped in 1990 ŽLeminen, 1991. with hand-held oblique photographs from June 1996 ŽRoss, 1998. indicates that the delta has advanced about 240 " 10 m in 6 years. The mean grain size of samples from core 95C1 in Silt Trap I, ranged from 9.2 to 47.8 mm Žmean, 20.1 mm., much higher than in cores from more distal regions of the reservoir. The structure of the core was almost entirely massive with a few silty layers, which made it impossible to correlate rates of sediment accumulation with those of Sthapit and Leminen Ž1992.. Because 95C1 was sub-sampled in the field, it was not possible to relate it to other

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323

314

Fig. 5. Sedimentary environments in Phewa Tal based on rates and processes of accumulation.

samples using MS as reported below with other cores. 5.2. Delta-proximal enÕironment The delta-proximal environment includes a portion of the former floodplain near the delta and much of the former river channel ŽFig. 5.. This region also receives significant input from the Harpan Khola. Cores 95C3, 95C4, 95C5 and 96C1 contain welllaminated, regular rhythmites best preserved in cores from the former floodplain; in some cases the rhythmites could be correlated between cores, especially in the upper 0.5 m. The lower half of each rhythmic couplet consists of from two to twelve graded silt beds Ž5Y 7r1. about 1- to 20-mm thick ŽFig. 6., and normally underlain by a dark-coloured, fine-grained layer less than 1-mm thick. Fine-grained clay caps commonly overlie the first few graded beds of these

units Že.g., at solid arrow, Fig. 6.. Finer-grained, massive deposits make up the upper halves of the couplets. These rhythmites are interpreted as varves, with graded beds from monsoon deposition, and massive beds from deposition of suspended fines during the dry season. Verification of the annual nature through 210 Pb dating of core 95C5 was attempted but because of the very high and variable rates of sedimentation, the 210 Pb profile was erratic, and could not be used to establish reliable dates. Other evidence exists of annual deposition. About 10 graded beds occur in the lower half of each couplet. Ten is also the average number of rainfall events per year with 24-h totals of 80 mm or more, which is the lower limit that caused noticeable increases in the turbidity of the Harpan Khola ŽRowbotham, 1995.. In the Jhikhu Khola watershed, Carver and Schreier Ž1995. reported that typically 15

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323

315

Fig. 6. Thin section of graded beds between 81 and 99 mm in core 95C3 showing typical patterns observed in monsoon and dry-season deposits. Solid arrow refers to discussion in text; open arrow indicates a crack in the thin section.

measurable erosion events occur per year. We propose that the graded beds in the lower half of each couplet represent storm events of this order, although the intensity necessary to generate distinct deposits

would vary because of seasonal changes in the availability of sediment and susceptibility to erosion as discussed above. This is analogous to the laminae in the summer deposits of glacilacustrine varves at-

316

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323

tributed by Gilbert Ž1975. and others to periods of accelerated melt of glaciers and to summer rain storms. The clay caps on the first few graded beds and the basal dark clay layers may represent initial mobilisation of organic matter from recently cleared agricultural fields and river banks by early monsoon rains. Little rain in the previous seven months, combined with a spring harvest, results in accumulation on the surface of organics and fine-grained material that are easily eroded. Where the lower clay layer is not observed, it is thought that significant early monsoon storms resulted in mobilisation of coarser-grained material and that the thin organic layer is not deposited because of higher energy levels in the fluvial and lacustrine systems. In such situations, clay caps are normally still present. Alternating light and dark layers in xradiographs of Phewa Tal cores ŽFig. 7. were also observed in cores from Nam Ngum, Laos where Axelsson Ž1992. attributed them to the high bulk density of monsoon deposits and low of dry season deposits, so that each lightrdark couplet is a varve. This was confirmed because the total number couplets matched the age of the reservoir. Further evidence of the annual nature of the Phewa Tal varves is seen in data of grain size, organic carbon, and carbonate ŽTable 1.. The laminated deposits in each couplet are consistently coarser than the massive sediments above, indicating erosion and transport in the higher-energy monsoon environment. The lower organic carbon and higher carbonate contents suggest that they represent the erosion of mineral soil which would be much more common during the heavy rains of the monsoon. In the cores, exceptionally thick ‘event’ deposits occur. For example, in 95C3, one of these from 0.269 to 0.346 m consists of a 15-mm thick massive silt bed overlain by a dark, finer grained 19-mm thick silty clay layer. Overlying this is a second massive silt bed Ž6 mm. which is capped by a finer grained bed Ž14 mm. and seven graded silt beds Žeach about 2 mm.. Varve counts indicate that these deposits represent the same event, and the date corresponds with the failure of the Pardi Dam in 1975 which lowered the lake level by about 3 m. We suggest the silts at the base represent erosion from the exposed lake floor or from the steepened lower

Fig. 7. Positive xradiograph of core 95C5 from 0.743 to 0.884 m, revealing monsoon and dry-season deposits based on differences in bulk density. Light areas represent sediments of low bulk density Ždry season.. Scale on left is in centimeters.

reaches of the Harpan Khola, and deposition at the core site. The graded beds above are interpreted as annual deposits formed as the exposed sediments were eroded annually in the monsoon until the dam was reestablished and the lake flooded. A similar pattern in glacially dammed Ape Lake was described

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323 Table 1 Mean grain size ŽGS., mean organic carbon content ŽOC., and mean carbonate content ŽCC. in the laminated and massive portions of varves from the delta-proximal zone of Phewa Tal Core

Parameter

95C3

GS Žmm. OC Ž%. CC Ž%. GS Žmm. OC Ž%. CC Ž%.

95C4

No. samples

Massive

26 21 21 29 29 29

5.01 4.04 6.20 6.16 5.30 5.32

317

1985. on the nearby southern shores of the reservoir, rather than from the upper watershed through the Harpan Khola.

Laminated

5.3. Distal enÕironment 6.50 3.42 6.57 7.41 3.78 5.84

by Gilbert and Desloges Ž1987. as a result of the failure of the glacial dam. Cores Žincluding the longest 95C4 at 2.30 m and 95C5 at 2.53 m. retrieved from more distal regions of the former river channel and floodplain in the delta-proximal environment also contain varves but these are fainter than those described above. Yearby-year chronologies cannot be determined. It is possible that some monsoon storms recorded elsewhere in the lake may not have sufficient energy to transport coarse material to this region. Rhythmic deposition is interrupted by a number of significant event deposits of poorly sorted, graded, 10- to 50-mm thick, sandy silt ŽRoss, 1998.. In these layers organic matter and carbonate content increase upward and each has elevated MS values. These characteristics suggest that they originated in mass movements on the relatively steep slopes of up to 408 ŽRamsay,

The distal environment of the reservoir ŽFig. 5. is dominated by massive sediments, deposited beyond the range of the interflows described above. Core 95C6 is typical of this region with the only observable change being its colour: from the surface to 0.67 m, massive fine silts darken from 5Y 4r2 to 5Y 2r2; from 0.67 to 0.90 m massive, lighter coloured Ž7.5YR 7r0. silts dominate. Disturbed darker laminae containing more organic material are found in the upper 0.11 m of the silt unit, suggesting a gradual transition between the depositional environments. Massive sediments are also found in the distal and shallow regions of some glacial lakes in which varves dominate the deep region Že.g., Gilbert and Desloges, 1987.. The darker sediment is associated with higher content of organic matter, whereas the lighter silts below are rich in carbonate ŽFig. 8.. The MS of the organic-rich sediment is significantly less. The origin of the carbonate-rich sediment is unknown but it may be associated with construction in Pokhara, possibly of the royal palace ŽRatna Mandir. located on the shore nearby ŽFig. 1.. Carbonate-rich sediment from the Seti Khola valley is regularly used as building material in Pokhara.

Fig. 8. Grain size, organic carbon content, carbonate content, and magnetic susceptibility in core 95C6.

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323

318

A similar pattern to that seen in the upper 0.69 m of 95C6 is observed in cores 96SC5 and 96SC8. No laminations are present but a change in colour from light Ž5Y 5r2. to dark Ž5Y 3r1. is observed. Similarly, the sediments of core 96SC5 are massive because of its location on the former floodplain at a depth of about 5 m. As discussed above, most major interflows are observed between 8 and 12 m depth, with minor interflows sometimes observed between 4 and 6 m. The lack of structure indicates that individual flows from even major events do not reach this portion of the reservoir. The overall darkening trend with depth observed in these cores is similar to that seen in cores 95C3 and 96C1, although it occurs at lesser depths in the cores from the distal environment. This reflects the lower rates of sedimentation in these regions. It may also indicate that these areas on the former floodplain were submerged more recently than deeper regions of the reservoir, as larger dams were constructed. 5.4. EnÕironment near the dam Attempts to retrieve sediment samples from the lake floor near Pardi Dam yielded only loose gravels with little mud. Sediment deposited in this region

under calm conditions in the dry season is likely eroded in high flows during the monsoon when the sluices of the dam are open. Although sediment delivery to the reservoir during the monsoon is greatest, the turbulence associated with the discharge of water from the dam at this time keeps much of the transported material in suspension, and, therefore, limits sediment accumulation in this region. 6. Sedimentary environment of Phewa Tal since 1900 The poor correlation between sediment cores from Phewa Tal precludes construction of reliable chronologies for each core based on stratigraphy alone. However, the varve dating record from core 95C3 combined with similar MS profiles from other cores ŽFig. 9. allows assessment of the evolution of erosional and sedimentary environments in this watershed. The correlation shown in Fig. 9 is based on consistent correspondence of peaks and troughs in the respective records. MS in each core rises from the bottom of the cores to a peak at about 0.76 to 0.86 m, followed by a subsequent decrease in MS to the surface. This pattern indicates that from about 1921 Žthe base of core 95C3. to 1955, the sediments being transported

Fig. 9. Magnetic susceptibility Ždimensionless. in cores 95C3, 95C4 and 95C5.

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323

to the reservoir were progressively richer in magnetic minerals. Since then, magnetic minerals show a progressive decrease to the present. This corresponds to a time of agricultural expansion and population growth in the watershed. Fires enhance the magnetic properties of soils devoid of primary ferrimagnetic minerals because of the reduction of non-ferrimagnetic oxides and oxyhydroxides to ferrimagnetic magnetite at extremely high temperatures ŽDearing et al., 1985.. Slash and burn techniques, common in the watershed in the early twentieth century until 1957 ŽSchroeder, 1985., may account for some of the enhanced MS in the core records. Magnetic properties of soils are also enhanced by pedogenesis ŽDearing et al., 1985.. The erosion of well-developed soil following reduction of forest cover in the watershed and the transportation of this

319

material into Phewa Tal also likely contributed to elevated MS values from the 1920s to the 1950s. Declining values since reflect transportation of lessdeveloped and, thus, less magnetically enhanced soils which are continually exposed by intense tillage. The high-resolution sedimentary record in core 95C3 reveals changes in erosional patterns resulting from the expansion of agricultural activity in the watershed ŽFig. 10.. Varve thickness and the relative thickness of monsoon vs. dry-season deposits in core 95C3 change after 1955 ŽFig. 10.. Mean thickness increases slightly from 17.5 to 18.8 mm Žnot including the three anomalously thick couplets at 0.269, 0.346 and 0.682 m.. The ratio of the thickness of monsoon to dry-season deposits ŽFig. 10b. also increases at this time, with a subsequent decrease since about 1980.

Fig. 10. Ža. Varve thickness and Žb. ratio of the thickness of monsoon deposit to dry season deposit for part of core 95C3. The period from 1975–1980 is accounted for by the deposit representing the failure of the Pardi Dam Ž0.269 to 0.346 m..

320

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323

The widespread removal of vegetation, the expansion of agriculture in the early to mid 1900s, and the intensification of farming Žsingle- to triple-cropping, relay cropping and inter-cropping: Schroeder, 1985. since this time likely resulted in enhanced early monsoon and full monsoon erosion. In turn, this has led to increased rates of sedimentation in the reservoir during this season. The subsequent decrease in the relative thickness of these deposits since about 1980 may reflect the stabilizing effect of well-maintained agricultural terraces, and an improvement in the techniques of land-management in recent years. A goal of this research was to determine if the sediments of Phewa Tal would provide annual records that could be used to calculate rates of sedimentation throughout the reservoir and determine its life expectancy. Although varves were observed in many cores, the sedimentary record is ambiguous and 210 Pb dating yielded inconsistent results. At best, an approximation of recent rates of sediment accumulation can be made for the Main Reservoir region of the lake based on the measurement of individual varves from 95C3 and 96C1 and the correlation of the chronology of 95C3 Žthrough MS and event deposits. with 95C4 and 95C5 in the former river channel. The average annual rate of sedimentation on the delta-proximal former floodplain, as recorded by 95C3 and 96C1 from 1955, is 22.0 mm ay1 . The peak in MS in 95C3 is found at 0.765 m. The same peak is observed in 95C4 at 0.795 m, indicating a slightly higher rate of accumulation in this region of the reservoir Ž23.4 mm ay1 .. Although disturbance of the core is noted at this depth, the offset of the MS profiles of both cores is consistent with depth, suggesting that this is a reasonable correlation. In 95C5, the same peak is observed at 0.855 m, which indicates a sediment accumulation rate of 25.1 mm ay1 . This is consistent with the location of event deposits slightly deeper in this core than in 95C3, 95C4 and 96C1 as discussed above, and supports the idea of higher rates of sedimentation on the slope of the lake-floor between Barahi Island and the southern shore. Using these estimates of annual sedimentation, the average annual rate of sediment accumulation in the Main Reservoir is 23.5 mm ay1 which represents a total of 78.7 = 10 3 m3 ay1 or 107 = 10 3 t ay1

based on an average wet bulk density of 1360 kg my3 . The result corresponds well with the calculation of Impat Ž1981. at 117 = 10 3 t ay1 based on suspended sediment concentrations of the Harpan Khola over a 12-month period. It is also similar to the estimate of Sthapit Ž1995. of sediment delivery to the reservoir at 175 to 225 = 10 3 m3 ay1 , with 90 to 120 = 10 3 m3 deposited in the Silt Traps, and 55 to 135 = 10 3 m3 deposited within the Main Reservoir. The volume of the Main Reservoir is approximately 35.5 = 10 6 m3. Based on these sedimentation rates, the life expectancy of this portion of the reservoir Žthe time at which 80% of storage capacity is lost. is approximately 360 years. Because the Main Reservoir contains 94% of the total volume of Phewa Tal, this estimate can be used as a guide for the life expectancy of the reservoir as a whole. This approximation is based on current rates and patterns of sedimentation, and once the storage capacity of the Silt Traps is lost Žestimated by Sthapit, 1995, to be within 20 to 25 years., sediment load to the Main Reservoir will likely increase considerably.

7. Conclusions Fig. 11 illustrates the influences on the sedimentary environment of the Phewa Tal watershed which have resulted in the spatial variability of the sedimentary record in the reservoir. This model identifies the allochthonous controls on sediment delivery to Phewa Tal, and autochthonous controls on sediment accumulation. Highly seasonal processes associated with the monsoon result in an annually fluctuating input of sediment to the lake over two orders of magnitude. Naturally high rates of erosion associated with the monsoon and with steep slopes of the easily eroded bedrock in the basin, have been augmented by human activity in this century. Thus, the sedimentary record of Phewa Tal is a complex response to natural and human processes, acting within and beyond the lake. Nevertheless, many of the sedimentary processes and deposits of monsoon lakes are analogous to those of glacial lakes, as seen in the physical properties and stratigra-

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323

321

Fig. 11. Factors influencing sediment delivery from the watershed, and sediment dispersal and accumulation in Phewa Tal.

phy of the varves. Knowledge of glacilacustrine sedimentology is usefully applied in the monsoon environment. Lacustrine sediments of lakes such as Phewa Tal have the potential to provide an enormous amount of information about geomorphic and human processes in Middle Mountain watersheds. This information can be used to augment studies of current watershed-scale processes by providing an important historical context. It also contributes to understanding the debate ŽIves and Messerli, 1989. regarding the extent of human influence on geomorphic processes in this region. Other proxy indicators, such as pollen or diatoms, can enhance the study of the

sedimentary record to facilitate inter-core correlation and the development of basin-wide chronologies.

Acknowledgements Research was carried out with a grant from the Natural Sciences and Engineering Research Council of Canada. S. Adhikary and C.K. Sharma in their official capacities with the Himalayan Climate Centre, Kathmandu, provided vital support and valuable input for the field operation. Formulation of the work was greatly aided by discussion with D.N. Row-

322

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323

botham and J.R. Jones, while P.B. Nepali’s daily assistance on site made the project possible.

References Anonymous, 1980. Management plan for the integrated development of Phewa Watershed, 1980–1985 Ž2037–2042.. Phewa Tal Technical Report No. 12. Axelsson, V., 1992. Sedimentation in the Nam Ngum Reservoir, Lao PDR. A Report Submitted to the Mekong Secretariat. AB Hydroconsult, Sweden, 69 pp. Carson, B., 1985. Erosion and Sedimentation Processes in the Nepalese Himalaya. International Centre for Integrated Mountain Development ŽICIMOD. Occasional Paper No. 1. ICIMOD, Kathmandu, 39 pp. Carver, M., 1997. Diagnosis of Headwater Sediment Dynamics in Nepal’s Middle Mountains: Implications for Land Management. Unpublished PhD Thesis. University of British Columbia, Vancouver, British Columbia, 374 pp. Carver, M., Nakarmi, G., 1995. The effect of surface conditions on soil erosion and stream suspended sediments. In: Schreier, H., Shah, P.B., Brown, S. ŽEds.., Challenges in Mountain Resource Management—Processes, Trends and Dynamics in Middle Mountain Watersheds. Proceedings of a Workshop Held in Kathmandu, Nepal, 10–12 April, 1995, pp. 155–162. Carver, M., Schreier, H., 1995. Scale influence on water and sediment output in a first-order mountain basin in Nepal. In: Osterkamp, W.R. ŽEd.., Effects of Scale on Interpretation and Management of Sediment and Water Quality. International Association for Scientific Hydrology, Wallingford, Oxfordshire, UK Publication No. 226, pp. 3–9. Dean, W.E. Jr., 1974. Determination of carbonate and organic matter in calcareous sediments and sedimentary rocks by loss on ignition: comparison with other methods. Journal of Sedimentary Petrology 44, 242–248. Dearing, J.A., Maher, B.A., Oldfield, F., 1985. Geomorphological linkages between soils and sediments: the role of magnetic measurements. In: Richards, K.S., Arnett, R.R., Ellis, S. ŽEds.., Geomorphology and Soils. George Allen and Unwin, London, pp. 245–266. Decore Consultancy Group, 1991. Socio-economic Baseline Survey of Phewa Tal Watershed. Department of Soil Conservation and Watershed Management, Ministry of Forests, Kathmandu. Desloges, J.R., Gilbert, R., 1994. The record of extreme hydrological and geomorphological events inferred from glaciolacustrine sediments. In: Olive, L.J., Loughran, R.J., Kesby, J.A. ŽEds.., Variability in Stream Erosion and Sediment Transport, International Association of Hydrological Sciences Publication No. 2, Vol. 224, pp. 133–142. Desloges, J.R., Gilbert, R., 1998. Sedimentation in Chilko Lake: a record of the geomorphic environment of the eastern coast mountains of British Columbia, Canada. Geomorphology 25, 75–91. Fleming, W.M. 1978. Phewa Tal Catchment Management Pro-

gramme. Nepr74r020 Project Working Document. Department of Soil and Water Conservation and Integrated Watershed Management Project. Food and Agriculture Organization of the United Nations, 110 pp. Fort, M., Freytet, P., 1982. The Quaternary sedimentary evolution of the intra-montane basin of Pokhara in relation to the Himalaya Midlands and their hinterland. In: Singh, A. ŽEd.., Contemporary Geoscientific Research in the Himalaya, Vol. 2, pp. 91–96. Gilbert, R., 1975. Sedimentation in Lillooet Lake, British Columbia. Canadian Journal of Earth Sciences 12, 1697–1711. Gilbert, R., Desloges, J.R., 1987. Sediments of ice-dammed, self-draining Ape Lake, British Columbia. Canadian Journal of Earth Sciences 24, 1735–1747. Gilbert, R., Glew, J., 1985. A portable percussion coring device for lacustrine and marine sediments. Journal of Sedimentary Petrology 55, 607–608. Gilbert, R., Desloges, J.R., Clague, J.J., 1997. The glacilacustrine sedimentary environment of Bowser Lake in the northern Coast Mountains of British Columbia, Canada. Journal of Paleolimnology 17, 331–346. Hickel, B., 1973. Limnological investigations in lakes of the Pokhara Valley, Nepal. Int. Revue ges. Hydrobiol. 58, 659– 672. Impat, P., 1980. Phewa Tal Watershed Soil Map. Integrated Watershed Management Project, Nepal. Impat, P., 1981. Hydrometeorology and Sediment Data for Phewa Watershed: 1979 Data. Ministry of Forest and Soil Conservation, Department of Soil Conservation and Watershed Management, Project Field Document No. 15, 29 pp. Ives, J.D., Messerli, B., 1989. The Himalayan Dilemma: Reconciling Development and Conservation. Routledge, London. Konrad, J.G., Chesters, G., Keeney, D.R., 1970. Determination of organic- and carbonate-carbon in freshwater lake sediments by a microcombustion procedure. Journal of Thermal Analysis 2, 199–208. Lamoureux, S.F., 1994. Embedding unfrozen lake sediments for thin section preparation. Journal of Paleolimnology 10, 141– 146. Leminen, K., 1991. Phewa Tal Siltation Survey—Base-line Data and the First Results 1990–1991. Integrated Watershed Management Project Technical Paper No. 2r1991, 43 pp. Leonard, E.M., 1986. Use of lacustrine sedimentary sequences as indicators of Holocene glacial history, Banff National Park, Alberta, Canada. Quaternary Research 25, 199–214. Lohmann, K., Jones, J.R., Knowlton, M.F., Swar, D.B., Pamperl, M.A., Brazos, B.J., 1988. Pre-and postmonsoon limnological characteristics of lakes in the Pokhara and Kathmandu Valleys, Nepal. Verhandelingen Internationale Vereinigung fur Limnologie 23, 558–565. Overseas Development Agency, 1995. Land Use, Soil Conservation and Water Resource Management in the Nepal Middle Hills. UK Institute of Hydrology, Royal Geographic Society, 12 pp. Pickrill, R.A., Irwin, J., 1983. Sedimentation in a deep glacier-fed lake—Lake Tekapo, New Zealand. Sedimentology 30, 63–75. Ramsay, W.J.H., 1985. Erosion in the Middle Himalaya, Nepal

J. Ross, R. Gilbertr Geomorphology 27 (1999) 307–323 with a Case Study of the Phewa Valley. Unpublished MS Thesis. University of British Columbia, Vancouver, British Columbia, 160 pp. Rana, K., 1990. Water Pollution Assessment of Phewa Lake, Pokhara, Nepal. Unpublished MS Thesis. Asian Institute of Technology, Bangkok, Thailand, 105 pp. Research and Soil Conservation Sections, 1994a. Sedimentation Survey of Phewa Lake—January, 1994. Ministry of Forest and Soil Conservation, Department of Soil Conservation, 26 pp. Research and Soil Conservation Sections, 1994b. Sedimentation Survey of Kulekhani Reservoir, October 1994. Ministry of Forest and Soil Conservation, Department of Soil Conservation, 17 pp. Ross, J.D., 1998. Erosion and Sedimentation in the Phewa Tal Watershed, Middle Mountain Region, Nepal. Unpublished MS Thesis, Queen’s University, Kingston Ontario, 174 pp. Rowbotham, D.N., 1995. Applying a GIS to the Modelling of Slope Stability in Phewa Tal Watershed, Nepal. Unpublished PhD Thesis, University of Waterloo, Waterloo, Ontario, 160 pp. Schreier, H., Shah, P.B., Lavkulich, L.M., Brown, S., 1994. Maintaining soil fertility under increasing land use pressure in the Middle Mountains of Nepal. Soil Use and Management 10, 137–142. Schroeder, R.F., 1985. Himalayan subsistence systems: indigenous agriculture in rural Nepal. Mountain Research and Development 5, 31–44.

323

Science Applications International, 1993. Environmental Protection Study of Phewa Lake, Pokhara Project Nepal. SAIC, VA, 252 pp. Shankar, K., 1991. Status and Role of Mountain Hydrology in the Hindu Kush–Himalayan Region. Mountain Environmental Management Discussion Paper No. 10. International Centre for Integrated Mountain Development ŽICIMOD., Kathmandu, 34 pp. Sharma, C.K., 1980. Natural Resources of the Pokhara Valley. Mrs. Sangeeta Sharma, Kathmandu, 106 pp. Sthapit, K.M., 1995. Sedimentation of lakes and reservoirs with special reference to the Kulekhani reservoir. In: Schreier, H., Shah, P.B., Brown, S. ŽEds.., Challenges in Mountain Resource Management—Processes, Trends and Dynamics in Middle Mountain Watersheds. Proceedings of a Workshop Held in Kathmandu, Nepal, 10–12 April, 1995, pp. 5–12. Sthapit, K.M., Leminen, K., 1992. Siltation Survey of Phewa Lake—Annual report of 1992. Integrated Watershed Management Project Technical Paper No. 3r1992, 24 pp. Thapa, G.P., Weber, K.E., 1995. Status and management of watersheds in the Upper Pokhara Valley, Nepal. Environmental Management 19, 497–513. Yamanaka, H., Yoshida, M., Arrita, K., 1982. Terrace landform and Quaternary deposit around Pokhara Valley, Central Nepal. Journal of the Nepal Geological Society 2, 113–142, Special Issue.