Soils as proxies of environmental fluctuations at the southern slopes of the Bhutan Himalayas
This study was conducted over a period of three years (2012 to 2015) with funding support from the German Ministry of Education and Science. The study focused on using soils as proxies to reconstruct the history of paleoclimate and landscape formation during the Quaternary Period especially of the Holocene Epoch. While there is a reasonably good record or account of Quaternary climate and landscape formation under the influence of South Asian Summer Monsoon (SASM) for various regions of the Himalayas, information on the Eastern region of the Himalayas, particularly for the Bhutan Himalayas is scarce and sketchy. Therefore, this study attempted to see how reliable physicochemical properties of soils can be interpreted to reconstruct Holocene and Pleistocene environment fluctuations, as soil formation and landscape development are impacted by the fluctuating climate, in particular by the SASM in the region. SASM is reported to have significant effects on the nature of glaciation throughout the Himalayas (Benn and Owen, 1998), and on soil formation and landscape development further downstream. Hence the results from this study try to relate (Holocene) summer monsoon intensities, glacial extents (moraine deposits) and their impact on the formation of soils and landscapes in two places of the Bhutan Himalayas. Using soils to reconstruct past climatic conditions is an important tool to test i) the reliability and applicability of soils as proxies in paleoclimate studies, and ii) to reconstruct the history of environmental fluctuations. The two study sites included Paro Dzongkhag in the western region and Bumthang Dzongkhag in the east central region of the country. Both valleys are traversed by perennial rivers, the Pa Chhu in Paro and Chamkhar Chhu in Bumthang, and both rivers are of glaciers origin, which are mainly fed by summer monsoonal precipitation. A total of 64 soil profiles were examined and 357 soil samples were collected from these two places. The altitudes from which the samples were collected ranged from 2,191 to 4,208 m asl. in Paro, while in Bumthang it was between 2,570 and 3,772 m asl. Soil chemical and physical parameters of samples from each horizon of the 64 soil profiles were analysed at the University of Erlangen and Nuremberg, Germany. The methodical approach included particle size analysis, the analysis of pedogenic iron oxides i.e silicate bond iron (Fet-d), well crystallized iron oxides (Fed-o), poorly crystallized oxidic iron compounds (Feo), and the calculation of eight weathering indices from total element contents i.e. Chemical Index of Alteration (CIA), Plagioclase Index of Alteration (PIA), Chemical Index of Weathering (CIW), Chemical Proxy of Alteration (CPA), Weathering Index of Parker (WIP), Index B, Silicon-Aluminium Ratio Index (SA) and Silica-Titania (STI). A time related soil development graph was established based on the data set of Bäumler (2001), and solum weighted means of the analysed soil parameters along with numerical ages determined by radiocarbon and Optically Stimulated Luminescence (OSL) dating were used to decipher the past environment information stored in the soils of the Bhutan Himalayas. The indicators used in this study were effective in determining the relative ages of the soils and therefore in estimating the impacts of environmental conditions on soil formation, landscape development and soil productivity. The results show increasing soil age with increasing terrace heights above the recent river levels, while the relation between soil age and altitude is not yet clear and needs further studies. Soil depth increases with increasing soil age as well. The proportions of finer soil particles (i.e. silt and clay fractions) increase with the extent of soil weathering. Similarly, the proportion of well crystallized iron oxides (Fed-o) is higher in more weathered soils as compared to either the silicate bond iron (Fet-d) or poorly crystallized iron oxides (Feo). The value of the ratio of Fed-o/Fet is also higher for older soils compared to younger soils. The contents of different elements in the soil indicate enrichment of elements with low mobility and low solubility and depletion of those elements which are highly soluble and mobile in older soils. The enrichment and depletion of elements with different mobility or solubility can be clearly seen at both sites of Paro and Bumthang. The general trend is declining nutrient contents (e.g. K2O) with increasing weathering or soil age. The soil particle size fractions, pedogenic iron oxides, element contents, Fed-o/Fet correlate well with the calculated ages of the soils. Some of the soil profiles such as PR21 and PR24 from Paro or PR37 from Bumthang are indicative of having non-volcanic andic soil properties. Accumulation of high amounts of iron oxides in the form of poorly crystallized oxidic iron compounds together with conditions suitable for Podzols, but without visible signs of podsolization, is a typical characteristic of non-volcanic andic soil. Soils with such characteristics have been described as non-volcanic soils with andic properties (Bäumler and Zech, 1994; Bäumler et al., 2005; Caspari, 2007). They are characterized by in-situ weathering and mineral transformation, and translocation in a leaching environment. Recent information indicates such soils to be widespread between altitudes of about 2,200 and 3,500 m asl., spanning over several bioclimatic zones. However, in this study, the fact that the profile PR21 is located at an altitude of 3,687±6 m and PR24 at 4,169±16 m asl. the results indicate an even wider range of non-volcanic andic soils across the Bhutan Himalayas. Out of the eight weathering indices tested in this study, six indices correlate well with the calculated soil ages. SA and STI indices did not appear to be as reliable as the other six indices. The values of CIA, PIA, CIW and CPA increase while those of WIP, Index B, SA and STI decrease with increasing extent of soil development. The results show most of the Paro soils to be of Holocene origin formed within the last 10 to 11.5 kyr BP, and few are of late Pleistocene origin developed between 10 to 60 kyr BP except for PR14 from Bondey which is comparably old (ca. 145 kyr BP). PR14 is characterised by an aeolian cover of about 55,000 years, and a red subsoil of about 145,000 years of age. The aeolian cover is about 90 kyr younger than the subsoil which has been developed in-situ from phyllite. A clear distinction in the horizons’ colours could be observed as dark yellowish brown (10YR) coloured horizons of the aeolian cover on top of the yellowish red paleosoil (5YR), a typical soil colour for subsoils developed from intensively weathered parent materials. All physical and chemical properties indicate this site as a well-developed and the most intensively weathered soil of all studied soils from both sites. The huge difference in the ages of the PR14 top and subsoil compared to all other soils made it very helpful for comparing the chemical and physical processes in soil weathering. Bumthang soils are older with estimated ages ranging from ca. 5.5 kyr BP to 67 kyr BP, and the majority is of late Pleistocene time. The ages of the younger soils correspond well with the time of Holocene SASM fluctuations and corresponding glacial activities reported by a number of studies. Evidences of numerous glacial advance and retreat cycles since the late Pleistocene in the far northern reaches of Bhutan were reported (Iwata et al., 2002; Meyer et al., 2009). Himalayan lake records indicate strong monsoonal activities around 25 kyr, 23.5 kyr, 22.5 kyr, 22-18 kyr, 17-16.5 kyr and after 14.5-13 kyr with strong early Holocene monsoon from 11.5 to 5.5 kyr (Kale et al., 2004; Patnaik et al., 2012). Warm and wet regional climate under the influence of strong SASM was also reported between 60 kyr and 30 kyr BP (Phillips et al., 2000). The reports on the glacial activity during the quaternary period indicate strong correlation with the intensities of SASM. Much of the glacial expansions reported occurred sometime close to the LGM at around 18 to 24 kyr BP in this region of the Himalayas (Owen et al., 2002). Studies conducted in the Kanchenjunga Himal (closer to Bhutan) reported glacial advances at around 6, 9, and 22 to 23 kyr (Asahi and Watanabe, 2000), and glaciers at Nanga Parbat in western Himalayas expanded at around 9.0 to 5.5 kyr (Phillips et al., 2000). The valley glaciers in Bhutan expanded during three distinct periods i.e. (1) the Little Ice Age and late Holocene glacial advances, (2) the Late Glacial and/or early Holocene glacial advances, and (3) the Last Glacial period (Iwata et al., 2002; Iwata, 2010). The results indicate glacier expansion in Paro and Bumthang to have taken place earlier than the Little Ice Age (ca. 600 to 200 yrs.) or Late Holocene glacial advances. This indicates that the intensity of glacial advancement and therefore the extent of moraine deposits was greater during the mid-Holocene than soon after the end of the LGM or late Pleistocene in both valleys. The glacial expansion in the Paro and Bumthang valleys can be assumed to have been broadly synchronous, not just between the two valleys but also with places like Kanchenjunga Himal (Owen et al., 2002), Khumbu Himal (Iwata et al., 2002) in the eastern region and Nanga Pharbat (Phillips et al., 2000) in western Himalaya. The soils of the fluvial terraces down-valley at both sites appear to have been deposited largely during the periods of high precipitation and warmer temperatures of Holocene (MIS-1) from 11.6 kyr BP onwards, and during the late Pleistocene (MIS-3) between ca. 60 and 24 kyr BP. Hence based on these time periods of fluvial depositions down the valleys, most of the Paro terraces can be said to be of Holocene origin while most of the Bumthang terraces are of late Pleistocene age. From this study, it can be concluded that various soil chemical and physical parameters and other indicators derived from them can be used effectively in determining the extent of soil weathering. Hence, the results show that soils can be used as proxies to reconstruct the history of landscapes and climate. Informations on local as well as global climate fluctuations are well preserved with pedogenesis. Besides indicating past climatic conditions under which soils are formed, soils and associated morphological features like terraces can be used to estimate or predict future landscape/landuse development and productivity of land under the influence of predicted climatic conditions. This study also indicates that the calculation of solum-weighted means of individual soil parameters is an effective tool to compare different sites and to minimize discrepancies due to the heterogeneity of parent materials of glacial, fluvioglacial, fluvial or aeolian origin.