The spatial and temporal distribution of fossil-fuel derived pollutants in the sediment record of Lake Baikal, eastern Siberia. *. N. L. Rose1, P. G. Appleby2, J. F. ...
Journal of Paleolimnology 20: 151–162, 1998. c 1998 Kluwer Academic Publishers. Printed in the Netherlands.
The spatial and temporal distribution of fossil-fuel derived pollutants in the sediment record of Lake Baikal, eastern Siberia N. L. Rose1 , P. G. Appleby2 , J. F. Boyle3 , A. W. Mackay1 & R. J. Flower1 1
Environmental Change Research Centre, University College London, 26 Bedford Way, London, WC1H 0AP, UK Department of Mathematical Sciences, University of Liverpool, P.O. Box 147, Liverpool, L69 3BX, UK 3 Department of Geography, University of Liverpool, P.O. Box 147, Liverpool, L69 3BX, UK 2
Received 2 December 1996; accepted 20 March 1997
Key words: atmospheric deposition, lake sediments, fly-ash particles, Lake Baikal
Abstract Spatial and temporal patterns of spheroidal carbonaceous particles (SCP) extracted from lake sediments provide an unambiguous record of the distributions of fossil-fuel derived pollutants. When applied to sediment cores taken from Lake Baikal spatial patterns show good agreement with the distribution of industry, with the highest concentrations found in the southern basin nearest to Irkutsk. SCP were found to occur in all cores from all areas of the lake in contrast to metal results where anthropogenically enhanced deposition was only demonstrable in the southern basin. SCP distribution within the sediments of Lake Baikal is seen to be distinctly regional and therefore long distance transport is not thought to be an important pathway for these pollutants. Temporal patterns of SCP show trends that reflect the development of industry in the area since the 1940s. Settling rates in the 1600 m water column suggest that the SCP sediment record may be approximately an order of magnitude more sensitive to depositional changes than that of trace metals. Introduction Lake Baikal is internationally famous for its rich flora and fauna with over 1 000 of the 2 500 identified animal and plant species believed to be endemic. The biological uniqueness of the lake stems from both its great age (current estimates vary between 25–50 million years) and size (over 1600 m deep and a volume of 23 000 km3 ) and the fact that its waters are oxygenated down to its greatest depths. Threats to the ecosystem of the lake have resulted from increased levels of pollution from industrial and domestic effluent, from atmospheric contamination and from logging in the catchment. Although these activities have been recognised throughout the twentieth century, concern has increased since the 1970s
This is the fourth in a series of seven papers published in this special issue dedicated to the paleolimnology of Lake Baikal. Dr. Roger Flower collected these papers.
and recently been brought to more global attention as a consequence of the break-up of the former Soviet Union (Stewart, 1990a&b). Effluent from partially treated sewage enters Baikal directly in the far north at Severobaikalsk and factory waste products enter the south of the lake both directly and via rivers, especially the Selenga River. However, the most notorious sources of pollution are the two pulp and cellulose mills, on the southern shore at Baikalsk and on the Selenga River at Selinginsk. Contamination of Lake Baikal via the atmosphere is a growing problem (Kokorin & Politov, 1991) and, due to recent increases in industrial emissions, may now pose a bigger threat to the ecosystem than point source water pollution (Stewart, 1990b). Sources of atmospheric pollution include not only the two pulp and cellulose mills but also other industries in the Irkutsk region (see Figure 1), which emit fly-ash, metals and sulphates to the atmosphere. However, the largest source of SO2 and particulates
152 Figure 1. Sediment coring sites in Lake Baikal together with the location of the main industries in the region. Size of circle corresponds to annual emission, subdivided to show the fraction relating to particulates and SO2 (Politov, pers. comm.). N.B. These data were compiled by Sergey Politov (Institute of Global Ecology and Climate Change, Moscow) from unpublished statistical records at the Regional Committee for the Protection of the Environment and Natural Resources, Irkutsk.
153 in the region is the oil-fired power station at Angarsk about 50 km downstream of Irkutsk on the Angara River. A number of smaller industries exist around the south of Baikal at Shelekhov, Usolye-Sibirskoe, Cheremkhovo, Sludyanka, Ulan-Ude, and Kamensk (Figure 1). A small amount of industry is also present in the north at Severobaikalsk and Nijneangarsk. Unlike effluents, whose effects are likely to be more localised, atmospheric pollutants can be carried long distances and their presence has been recorded in even the more remote regions of the Baikal area, for example the Khamar Daban mountains to the south-east (Flower et al., 1997). Lake Baikal supports the highest number of endemic species of any freshwater lake (Kozhov, 1963) and as a consequence it has been proposed as a World Heritage Site by UNESCO. It is therefore imperative to be able to evaluate changes in the Lake Baikal ecosystem and in depositional regimes impacting the lake and its catchment. The palaeolimnological record is able to provide evidence of temporal trends in ecological change and pollutant deposition as well as contemporary spatial patterns. Spheroidal carbonaceous particles (SCP) form an excellent sedimentary record of pollution emissions. They are produced by the high temperature combustion of fossil-fuels and as such are unambiguous indicators of anthropogenic impact from atmospheric deposition. Their temporal distribution as recorded in dated lake sediment cores agrees closely with records of fossilfuel combustion throughout Europe (Renberg & Wik, 1984; 1985; Wik & Renberg, 1996; Rose et al., 1995) and the USA (Charles et al., 1990). SCP spatial distribution has been shown to be closely linked with sulphur deposition (Wik & Renberg, 1991; Rose & Juggins, 1994) as well as with other pollutants such as polycyclic aromatic hydrocarbons (PAH) (Broman et al., 1990). SCP presence has been recorded in remote areas far from industrial sources such as Svalbard (Rose, 1995), Iceland and the Russian and Canadian Arctic (Rose, unpublished data) suggesting a possible hemispherical background level at these remote sites. This paper describes the spatial and temporal distribution of SCPs in Lake Baikal sediments, and from this evidence attempts to determine the extent of the impact from the various sources as well as the historical trends of atmospheric deposition of fossil-fuel derived pollutants.
Methods (i) Coring A single sediment core (BAIK 6) was taken in September 1992 from 1420 m depth using the Baikal box-corer (Flower et al., 1995a) and in July 1993 a further 29 sediment cores were taken using this box corer and a short gravity corer (Glew, 1991). The sediment retrieved in the box corer was sub-sampled using a wide diameter piston corer gently pushed into the sediment, minimizing smearing and compaction. Disturbance effects were further reduced by trimming each slice on extrusion. These sub-cores were sectioned immediately upon retrieval; the 0–5 cm section in 2 mm intervals, 5–10 cm in 5 mm intervals and the remainder in 10 mm intervals. The samples were sealed in plastic bags. Despite precautions, some of the sediment cores showed signs of surface disturbance and these were not analysed further. The locations of the sediment cores selected for SCP analysis are shown in Figure 1. (ii) SCP analysis The surface sediments from 26 cores were analysed for SCPs. In addition, the full SCP profile was determined for six 210 Pb dated cores (BAIK6, 19, 22, 25, 29 and 38). The SCP profile from BAIK6 has previously been published in Flower et al. (1995b). SCP analysis followed the method described in Rose (1994). Dried sediment was subjected to sequential chemical attack by mineral acids to remove unwanted fractions leaving carbonaceous material and a few persistent minerals. SCP are composed mostly of elemental carbon and although physically fragile are chemically robust. The use of concentrated nitric acid (to remove organic material), hydrofluoric acid (siliceous material) and hydrochloric acid (carbonates and bicarbonates) therefore does them no damage. A known fraction of the resulting suspension was evaporated onto a coverslip and mounted onto a microscope slide. The number of SCP on the coverslip were counted using a light microscope at 400 magnification and the sediment concentration calculated in units of ‘number of particles per gram dry mass of sediment’ (gDM,1 ). The detection limit for the technique is 100 gDM,1 and concentrations have an accuracy of 45 gDM,1 .
154 (iii) Sediment dating Radiometric dates were obtained for the sediment cores by measuring 210 Pb, 226 Ra, 137 Cs and 241 Am by gamma spectrometry (Appleby et al., 1986). 210 Pb is a naturally occurring radionuclide of half-life 22.26 years and measurements of the down-core decline in 210 Pb activity in excess of the supporting 226 Ra are used to determine a chronology for the past 100–150 years. 137 Cs and 241 Am are artificial radionuclides first introduced into the environment on a global scale in 1954 by the atmospheric testing of thermo-nuclear weapons. Fallout of 137 Cs from this source reached a maximum value in 1963 and then declined sharply following the treaty in that year banning further atmospheric tests. Sediment records of this maximum can be used to confirm the 210 Pb defined 1963 level in the core. Individual chronologies of the Baikal sediment cores are discussed in Appleby et al. (this volume).
Results and discussion (i) Spatial distribution The SCP concentrations from the 26 surface sediment samples (0–2 mm depth) distributed throughout the three basins of Lake Baikal are shown in Figure 2. There is a distinct pattern to these concentrations. The highest are in the southern basin with a maximum of 5 200 gDM,1 for BAIK39, the closest site to Irkutsk. A cluster of sites in the southern basin BAIK 8, 12, 33, 34 and 37 show concentrations greater than 3 000 gDM,1 and concentrations decrease away from Irkutsk both to the south of this area (BAIK 38) and to the north towards the middle basin. The middle basin shows the lowest concentrations and is the area furthest removed from industry. SCP concentrations in all surface sediments in this area are less than 1 000 gDM,1 (BAIK 20–25 inclusive) with higher concentrations in the extreme south (BAIK20) and north (BAIK25) of the middle basin. These low concentrations are similar to those found in other remote areas of the northern hemisphere (e.g. Spitsbergen, Canadian Arctic) and it has been suggested that such levels represent a hemispherical background concentration of SCPs in contemporary sediments (Rose, 1995). If this is the case, then this indicates that industrial emissions from Irkutsk in the south and other industry in the north are having little impact on the sediments of the central basin of Lake Baikal. Concentrations increase northwards in
the northern basin with the highest SCP concentration (2 200 gDM,1 ) in BAIK 28 the furthest north sediment core. This implies that atmospheric deposition (above any hemispherical background level) to the northern basin is from sources to the north of the lake. Sediment accumulation rates vary however and these can influence SCP concentrations making intercore comparisons less reliable. Recent sediment accumulation rates for the six 210 Pb dated cores vary from 0.017 (BAIK38) to 0.050 g cm,2 yr,1 (BAIK 22) but these can be corrected for, to a certain extent, by converting the SCP concentrations to SCP accumulation rates. The highest SCP accumulation rate is BAIK6 with 67 cm,2 yr,1 , and, like the concentration trends, these decrease away from Irkutsk to the north. However, whereas for concentrations the lowest values were in the middle basin and then increased again northwards, there is now less difference between the middle and northern basins. The lowest SCP accumulation rate is for BAIK 25 (5.7 g cm,2 yr,1 ), an order of magnitude lower than for BAIK6, but rates for BAIK22 to the south of BAIK25 and BAIK29 to the north are relatively similar at 16 and 14 cm,2 yr,1 respectively. Conversion to SCP accumulation rate therefore highlights the southern basin as the area of maximum particle deposition and pollutant impact suggesting that the Irkutsk region is the major source of industrial atmospheric emissions affecting the lake. Although BAIK28, the most remote coring site from Irkutsk, is over 460 km away it is not inconceivable that particulate emissions from Irkutsk could be transported over such distances. However, as the northern basin concentrations are higher than those in the middle basin, nearer to Irkutsk, and it is thought that these middle basin concentrations represent a hemispherical background, then it is probable that northern basin SCP concentrations are elevated above this background due to sources to the north of the lake, rather than due to long-distance transport from Irkutsk. These results are in good agreement with those reported by van Malderen et al. (1996). Aerosols collected over Lake Baikal in 1992/93 and analysed by EPXMA revealed the northern and middle basins to be similar and the southern basin to be most contaminated by particles of industrial origin. However, a significant percentage of particles collected over the north of the lake were found to be organic and these were thought to be from two sources, biogenic (pollen etc.) and from fossil-fuel combustion at Severobaikalsk. SCP analysis has also been undertaken on sediment cores from two small mountain lakes in the Khamar
Figure 2. SCP concentrations in the surface sediments shown as proportional circles. Concentrations in ‘number particles per gram dry mass of sediment’ (gDM,1 ).
156 Daban mountains south-east of Lake Baikal. Lake Kholodnoye and Lake Kvadratnoye were both cored in 1992 and a more detailed description of the results of the sediment analyses are given in Flower et al. (1994). The SCP surface concentration of Kholodnoye is 2 800 gDM,1 and that of Kvadratnoye is 2 550 gDM,1 . Both cores have been 210 Pb dated and surface SCP accumulation rates are 26.2 and 78.0 cm,2 yr,1 respectively. These SCP accumulation rates are in the same range as the cores in the southern basin of Baikal although the surface SCP accumulation rate for Kvadratnoye is higher than any of the dated Baikal cores. Kvadratnoye is nearer to Irkutsk (the assumed source) than Kholodnoye but more distant than the Baikal southern basin sites. However, these lakes are at altitude and it may be that there is some seeder/feeder enhancement to SCP deposition at these sites. Concentrations of pollutants have been reported as being five or six times more concentrated in cap cloud than in rain (Dore et al., 1992). Cloud droplets formed in cap clouds are too small to grow into raindrop-sized particles, but can be washed out when ‘seeder’ raindrops from higher level clouds fall through the ‘feeder’ cap cloud thereby producing elevated pollutant depositions at higher altitudes. This mechanism for enhancing SCP concentrations has been suggested as being possible at some lake sites in the U.K. (Rose & Juggins, 1994). The high SCP accumulation rates of BAIK6 and Kvadratnoye are still low in comparison to most European mountain lakes, especially in central Europe and the U.K. Sites in the Spanish Pyrenees studied as part of the EC funded research programme AL:PE (unpublished data) show similar or lower SCP accumulation rates (Laguna Aguilo – 51.0 cm,2 yr,1 ; Laguna Redo – 22.9) as do Lac Noir in the French Alps (63.0), and Laguna Caldera in the Sierra Nevada (54.3). In the U.K., only Loch Coire nan Arr in the north-west of Scotland shows a comparable contemporary SCP accumulation rate (29.4), all others being much higher. Sites in mid-Norway show SCP accumulation rates comparable to those in mid- and northern Baikal (Øvre Ne˚adalsvatn – 12 cm,2 yr,1 ) and the lowest Baikal SCP accumulation rate, BAIK25 (5.7) is of similar magnitude to that of the remotest European sites, for example, Arresjøen on Svalbard (1.3 cm,2 yr,1 ), again suggesting that the mid-Baikal region is receiving only background levels of pollutants.
(ii) Temporal distribution Figure 3 shows the SCP profiles for the six 210 Pb dated sediment cores and the temporal patterns are broadly similar to those seen in Europe i.e. there is, in general, a long period of low SCP concentration followed by a rapid increase to a maximum and a surface decline in three of the six cores (Rose et al., 1995). The longest SCP profile is for BAIK38 in the southern basin where the record appears to begin in the 1850s (Figure 3). There was little industrial activity in the area before 1940 (Politov, pers. comm.) and this early date could therefore be due to sediment smearing during coring or extrusion, despite the precautions outlined above. Small scale bioturbation or other physical disturbance causing the SCP rich sediment from upper levels to be moved down to lower sediment depths is also a possibility, although other measurements (see Appleby et al., this volume; Boyle et al., this volume) suggest that large scale mixing has not occurred. BAIK38 was taken from 600m depth compared to > 1000 m for the other cores and it may be that this core has been subjected to bioturbation to a greater sediment depth resulting in an apparently longer SCP profile. The start of the record in BAIK38 does not pre-date the usual start of the SCP record in Europe (1850s–1860s) and could possibly be caused by long distance transport of pollutants. However, the absence of particles in the other Baikal cores at this time makes this unlikely. The profile for BAIK19 shows a presence of SCP as early as 1905 (13 years), but this again is probably caused by smearing as the profile falls to a concentration of 0 gDM,1 above this (1940s) and the upper ‘zero’ should probably be considered the start of the SCP profile. BAIK6 and BAIK29 show the start of the particle record to be in the 1920s/1930s as does the profile from Lake Kvadratnoye (Flower et al., 1994). This suggests that 1920s/1930s is the most likely date for the start of the SCP sediment record in the Baikal region. This is in agreement with coal mining statistics for the Irkutsk area (Figure 4). This shows that the amount of coal mined increased rapidly from ca.1930 to the mid-1980s (Office of Statistics, Irkutsk District, pers. comm) (Figure 4). Almost all the coal mined in the Irkutsk area is burned locally (Grachev, pers comm.) and therefore should be a good surrogate for combustion statistics which are not available prior to 1954 (Irkutskenergo Co. pers. comm.). BAIK22 and BAIK25 show later profile starts, 1979 2 and mid-late 1960s respectively. However, concentrations in these short profiles are very low throughout suggest-
Figure 3. SCP profiles of the six 210 Pb dated sediment cores.
Figure 4. Historical statistics for mined coal (103 tons) and power station heat capacity (MW) for the Irkutsk area.
ing these late dates may be due to the limit of detection of the SCP technique. The start of the rapid increase in SCP concentration, usually a reliable indicator of the 1950s/1960s in the U.K. and Europe (Renberg & Wik, 1985; Rose et al., 1995), is less defined in the Baikal sediments. This feature is usually allocated to the depth at which the intercept of the two gradients (pre- and post-gradient change) occurs (Rose et al., 1995) and in the Baikal cores this varies from the mid-late 1960s (BAIK25) to the late-1980s (BAIK29), a period during which both coal mining and fossil-fuel combustion increased continuously in the Irkutsk area (Figure 4). The SCP rapid increase in BAIK29 is much later than for any of the sites in the southern basin and this is probably due to the more recent development and increase in emissions from industries in the north compared with those around Irkutsk. The low concentrations and short profiles for BAIK22 and BAIK25 make any assessment of SCP profiles rather difficult, and it may be that the start of the record in these cores represents the rapid increase in SCP concentration in other Baikal profiles. The situation for BAIK22 is further complicated by the presence of a turbidite layer at the beginning of the 1980s (Appleby et al., this volume) which may affect the bottom of the SCP profile in this core. Rapid sediment accumulation at this time could have caused
the SCP concentration to fall below the detection limit and hence the profile to show a zero concentration at a more recent date than would otherwise have been the case. A sub-surface SCP concentration peak followed by a surface decline is observed in three of the six cores (Figure 3), BAIK 19, 22 and 38. This decline is probably due to the reduction in fossil-fuel consumption (as indicated by coal mining statistics – Figure 4), as particle arresting equipment has not recently been introduced in the area. In two of the remaining cores, the absence of the peak can be explained by low concentrations and short profiles (BAIK25) and a lack of sample resolution for sediment levels from BAIK6. Where it exists the peak occurs at a similar date in all cores (1989 2 – 1991 2) and in this respect it is similar to the European SCP profiles where the SCP peak is the best defined of dating features (Rose et al., 1995). The remaining core, BAIK29, shows a surface maximum, but uniquely amongst the Baikal cores shows a slight sub-surface decrease just before it. The peak immediately before this decrease also dates to 1990 2 and therefore may represent the feature present in the other cores. If this is the case, then the SCP peak is the only feature to occur at the same time throughout Baikal and could therefore prove to be a useful sediment marker.
159 An alternative interpretation could be that the surface maximum in BAIK29 is due to the development of more local industry at the northern end of the lake whose impact is not observed in the southern basin. These interpretations rely on a single data point, the most recent, and so should be treated with caution. Future cores will need to be analysed to determine whether these are continuing trends or whether they are just artifacts of the sediment record. No surface declines are observed in the sediment metal record, although a similar pattern is observed for Zn, Pb and SCP in BAIK29. In the same way that SCP surface sediment concentrations can be converted to SCP accumulation rates to compensate for variations in sedimentation rate, so the SCP profiles can be converted to inventories covering the full period of deposition. As the full profile is accounted for in the SCP inventory, these allow for changes within each core and also allow comparisons between cores on a ‘total deposition’ basis. However, variations between sites still may exist due to, for example, the inwash of atmospheric pollutants from catchment sources and these can be compensated for by normalising the SCP inventories to the 210 Pb inventories for each core. The resulting SCP/210 Pb ratio is therefore a ‘pollution index’ for the deposition period and allows a better inter-site comparison for total SCP atmospheric deposition. These inventories are shown in Figure 5. The unmodified SCP inventories show similar values for the two southern basin sites (BAIK38 and BAIK6) and then a decrease northwards through BAIK19 to a consistent low value for BAIK 22, 25 & 29, suggesting little or no impact from the northern industries and that these values may be representing an SCP background figure. When normalised to 210 Pb inventories however, there are two main differences in the pattern. First, there is a more obvious south to north trend from BAIK 38 through to BAIK 22 in the middle basin. The BAIK 38 SCP/210 Pb ratio is now considerably higher than that of BAIK 6 and this may be due to its greater proximity to sources in Baikalsk, possibly the pulp and cellulose mill, whereas sources in Irkutsk and Angarsk may be more evenly distributed over the whole southern basin (i.e. approximately equal contributions to both BAIK 38 and BAIK 6). If this is the case, then this is significant evidence for the mill having an impact on Lake Baikal from its atmospheric emissions in addition to the impact caused by direct discharge of waste into the lake. Van Malderen et al. (1996) reported that the highest abundances of sul-
phur rich (> 80% S as measured by EPXMA) particles occurred in the atmosphere in the vicinity of the Baikalsk paper mill (6.4% abundance) and that lower but similar abundances occurred in the middle and northern basins (3.7% and 3.9% respectively). Second, the sites in the northern basin have slightly higher SCP/210 Pb ratios than the middle basin site and this slight difference may be due to the influence of the industries at Severobaikalsk and Nijneangarsk. The SCP/210 Pb inventory ratios for the southern basin (2–3 000) are of a similar order to European mountain lakes known to receive moderate pollution loads. For example, Stavsvatn in southern Norway (2 950), Lac Noir in the French Alps (2 800) and Lagoa Escura (Sierra da Estrela, Portugal) (2 600) show similar values (Rose & Appleby, unpublished data). The ratios for the middle (175) and northern basins (275– 375) of Baikal are much lower and only have European analogues in the cleanest and remotest sites (e.g. Arresjøen on Svalbard (125) and Øvre Ne˚adalsvatn (550) in mid-Norway). This supports the hypothesis that the middle basin is receiving only hemispherical background levels of SCP and the sites in the north are only slightly more contaminated presumably due to the industries to the north of the lake. The southern basin is seen to be reaching levels of atmospheric contamination which would be considered moderate in Europe. Comparisons with metals data The results obtained from the palaeolimnological investigations of Lake Baikal using both trace metals (Boyle et al., this volume) and SCPs are broadly similar. Metals only show enhanced supply in the southern basin with cores from the middle and northern basins showing no evidence for atmospheric deposition from anthropogenic sources. The SCP record which is, to a certain extent, less ambiguous than the metal record due to the fact that SCP are derived solely from atmospheric deposition of industrial emissions, also shows the southern basin to be the most contaminated. Metals as, or attached to, very fine particulates will have longer atmospheric residence times than the larger SCP and should therefore be able to travel longer distances before deposition. Consequently, a large source at the southern end of Baikal might be expected to show a more widespread distribution for metals than SCP but this is the opposite to that which is observed as analyses indicate a presence of SCP in all cores in all basins. The reason for this is that there
Figure 5. SCP inventories (105 particles m,2 ) and SCP/210 Pb inventory ratios for the six 210 Pb dated sediment cores.
161 is a high pre-industrial baseline for metals and consequently any anthropogenic signal is lost as distance from the southern basin is increased. The SCP in the northern basin are therefore most likely being produced from the industries in Severobaikalsk and Nijneangarsk although there are likely to be small contributions from industries in the south and possibly even long-distance transport from sources outside the region. SCP data in the northern and middle basins show little enhancement over and above a supposed hemispherical background figure suggesting impact from the northern industries is small and that any anthropogenic metals signal from these sources is being obscured by background noise. Other studies have shown the presence of aerosols containing heavy metals, most probably of anthropogenic origin to be found in all three basins (van Malderen et al., 1996), although abundances always constituted < 3% of the total aerosol. Both SCP and metals data for a small mountain lake in the Khamar Daban mountains to the south-east of Baikal (Flower et al., 1997) show good agreement with data from the southern basin cores indicating that the impact of atmospheric deposition from Irkutsk area sources is not confined to Lake Baikal but also to a considerable area around the south of the lake. Significant differences must exist between SCP and fine metal particulates with respect to the time taken for changes in depositional regime to appear in the sediment record. Boyle et al. (this volume) suggest that due to the extraordinarily long residence time of fine particles in the 1600m water column, the heavy metal record is insensitive to changes in deposition on a smaller than decadal time scale. Stokes’ Law states that spherical particles of a given density settle through water at a rate directly proportional to the square of their radii. SCP are large and composed mainly of elemental carbon. Theoretical calculation suggests that a SCP of radius 10 m would take in the order of 75 days to settle through a column of water 1600m deep. This is, of course, under ideal conditions with no resuspension, convection or current movement taken into account. Despite these however, it is likely that SCP will have a residence time in the Baikal water column an order of magnitude less than that of metals and likewise a comparable increase in response time to depositional changes in the sediment record. The sediment record of SCP in Lake Baikal therefore appears to be more sensitive to changes in atmospheric deposition than that of the metal record, especially on a short (< 10 year) time scale. Howev-
er, it has been suggested that for diatom frustules (of the same size order as SCP), temporary deposition on slopes can significantly increase the sediment response time (Grachev pers. comm.) and it maybe that SCP are subject to the same processes and that the SCP response time discussed above is a minimum. Since SCPs are only formed by high temperature combustion of fossil-fuels their only source is from atmospheric deposition (directly or indirectly). Consequently, there can be no ambiguity about their provenance. In addition, once deposited, SCP do not suffer from remobilisation due to chemical changes in the sediment that can sometimes disturb the metal record. It is therefore suggested that SCP provide a more faithful record of atmospheric deposition from anthropogenic sources than metals in Lake Baikal especially with regard to short term changes. Despite this, on a broad time-scale, changes in the sediment record of both SCP and metals are seen to occur at similar times (e.g. the Pb increase in the 1950s/1960s in the southern basin) and the conclusions drawn by Boyle et al. (this volume) from the metal record that there is good spatial correlation between ‘local’ pollution sources and distribution of pollutants in the sediments is supported by SCP evidence. However, whereas Boyle et al. suggest no anthropogenic enhancement to the metal record in the middle and northern basins, the SCP record shows low levels of contamination at all sites. This is probably due to the atmospheric metal signal being lost in background ‘noise’ whereas no comparable background for SCP exists and the presence of SCP is sufficient to indicate atmospheric deposition from industrial sources.
Acknowledgements We would like to thank Prof. M. Grachev, Anna Kuzmena, Dr Ye. Likhoshway and other members of the Limnological Institute in Irkutsk, the crew of the R.V. Titov and Don Monteith of the Environmental Change Research Centre, University College London for their support and help with the fieldwork. In addition we would like to thank Dr Grachev for his useful comments on the manuscript and for supplying the coal mining and heat capacity data used in Figure 4. We are grateful for the financial support received from the Royal Society (BICER), the Leverhulme Trust (Project F134AZ) and ENSIS Ltd. (University College London) enabling the work to be undertaken. Cath Pyke in the
162 Cartographic Office of the Department of Geography, University College London produced the figures.
References Appleby, P. G., P. Nolan, D. W. Gifford, M. J. Godfrey, F. Oldfield, N. J. Anderson & R. W. Battarbee, 1986. 210 Pb dating by low background gamma counting. Hydrobiologia 141: 21–27. Appleby, P. G., R. J. Flower, A. W. Mackay & N. L. Rose, this volume. Palaeolimnological assessment of recent pollution trends in Lake Baikal: Sediment chronology. Boyle, J. F., A. W. Mackay, N. L. Rose, R. J. Flower, P. G. Appleby & L. Z. Granina, 1998. Sediment heavy metal record in Lake Baikal: natural and anthropogenic sources. J. Paleolimnol. 20: 135–150. Broman, D., C. N¨af, M. Wik & I. Renberg, 1990. The importance of spheroidal carbonaceous particles for the distribution of particulate polycyclic aromatic hydrocarbons in an estuarine-like urban coastal water area. Chemosphere 21: 69–77. Charles, D. F., M. W. Binford, E. T. Furlong, R. A. Hites, M. J. Mitchell, S. A. Norton, F. Oldfield, M. J. Paterson, J. P. Smol, A. J. Uutala, J. R. White, D. R. Whitehead & R. J. Wise, 1990. Paleoecological investigation of recent lake acidification in the Adirondack Mountains, N.Y. J. Paleolimnol. 3: 195–241. Dore, A. J., T. W. Choularton & D. Fowler, 1992. An improved wet deposition map of the United Kingdom incorporating the seederfeeder effect over mountainous terrain. Atmos. Envir. 26A: 1375– 1381. Flower, R. J., S. Politov, N. Solovieva, N. L. Rose, R. W. Battarbee, B. Rippey, P. G. Appleby & C. L. Rose, 1994. A palaeolimnological evaluation of the extent and impact of atmospheric contamination and recent environmental change in two remote highland lakes in SE Siberia. Final report for the NERC small research grant GB9/1032. Environmental Change Research Centre, Research Report, No. 9. Flower, R. J., D. T. Monteith, A. W. Mackay, J. M. Chambers & P. G. Appleby, 1995a. The design and performance of a new box corer for collecting undisturbed samples of soft subaquatic sediments. J. Paleolimnol. 14: 101–111. Flower, R. J., A. W. Mackay, N. L. Rose, J. F. Boyle, J. A. Dearing, P. G. Appleby, A. E. Kuzmina & L. Z. Granina, 1995b. Sedi-
mentary records of recent environmental change in Lake Baikal, Siberia. Holocene 5: 323–327. Flower, R. J., S. Politov, B. Rippey, P. G. Appleby, N. L. Rose & A. C. Stevenson, 1997. The extent and impact of atmospheric contamination and recent environmental change in a remote highland lake in south-eastern Siberia. Holocene 7: 161–173. Glew, J. R., 1991. Miniature gravity corer for recovering short sediment cores. J. Paleolimnol. 5: 285–287. Kokorin, A. O. & S. V. Politov, 1991. Wet deposition in the south Baikal region. Soviet Meteorology and Hydrology 1: 39–44. Kozhov, M., 1963. Lake Baikal and its life. W. Junk. The Hague, Netherlands, 344 pp. Renberg, I. & M. Wik, 1984. Dating of recent lake sediments by soot particle counting. Verh. int. Ver. Limnol. 22: 712–718. Renberg, I. & M. Wik, 1985. Carbonaceous particles in lake sediments – pollutants from fossil-fuel combustion. Ambio. 14: 161– 163. Rose, N. L., 1994. A note on further refinements to a procedure for the extraction of carbonaceous fly-ash particles from lake sediments. J. Paleolimnol. 11: 201–204. Rose, N. L., 1995. Carbonaceous particle record in lake sediments from the Arctic and other remote areas of the Northern Hemisphere. Sci. Tot. Envir. 160/161: 487–496. Rose, N. L. & S. Juggins, 1994. A spatial relationship between carbonaceous particles in lake sediments and sulphur deposition. Atmos. Envir. 28: 177–183. Rose, N. L., S. Harlock, P. G. Appleby & R. W. Battarbee, 1995. Dating of recent lake sediments in the United Kingdom and Ireland using spheroidal carbonaceous particle (SCP) concentration profiles. Holocene 5: 328–335. Stewart, J. M., 1990a. Baikal’s hidden depths. New Scientist. 23 June: 42–45. Stewart, J. M., 1990b. ‘The Great Lake is in peril’. New Scientist. 30 June: 58–62. van Malderen, H., R. van Grieken, T. Khodzher, V. Obolkin & V. Potemkin, 1996. Composition of individual aerosol particles above Lake Baikal, Siberia. Atmos. Envir. 30: 1453–1465. Wik, M. & I. Renberg, 1991. Recent atmospheric deposition in Sweden of carbonaceous particles from fossil-fuel combustion surveyed using lake sediments. Ambio. 20: 289–292. Wik, M. & I. Renberg, 1996. Environmental records of carbonaceous fly-ash particles from fossil-fuel combustion. J. Paleolimnol. 15: 193–206.