Petrography and stable isotope geochemistry of Middle Eocene Garampani limestone (Assam), India: implications for the depositional environment and diagenesis

Petrography and stable isotopic (carbon and oxygen) studies of the Middle Eocene Garampani limestone from Jamunagar limestone quarry of the Umrangso area, North Cachar Assam, were carried out to determine the environment of deposition and the level of diagenesis to improve our understanding of the climate at the time of its formation. The major petrographic types documented here include wackestone, packstone, and grainstone which fall into the type I and II categories of the triangular classification scheme indicating formation in a short-lived, high-energy environment that was dissipating at the time of deposition. The samples have a distinct signature that includes extensive diagenesis with varying intensities of compaction, recrystallization, pressure solution, and neomorphism. In occurrences of stratigraphically significant foraminiferal assemblages and palynofossils, the Garampani Limestone is assigned Middle Eocene age. Most of the samples produced positive δ13C ‰ ranging from 1.64 to 0.76 ‰ (PDB) and correlated with a transgressive phase. The negative δ18O values range from −6.25 to −4.67 ‰ (PDB) suggesting that the Middle Eocene was generally characterized by a warmer climate.


Introduction
The Eocene Epoch (56-33.9 Ma) is a dynamic geological interval that contained both a period of ice-free greenhouse conditions, and the transition into an icehouse world. During this Epoch, among several short-lived hyperthermal events of the Paleocene-Eocene Thermal Maximum (PETM), early Eocene Climatic Optimum (EECO) and Middle Eocene Climatic Optimum (MECO) commenced with rapid warming events that released massive amounts of isotopic carbon into the ocean-atmosphere system, and were followed by a sustained period of warmth (Zachos et al. 2001;Patra et al. 2021). The Eocene epoch is also a period of mountain building in the Himalaya driven by, and the rapid northward drift of Indian plate and the resulting continental collision, followed by a complete closure of eastern Neo-Tethys (Klootwijk et al. 1992;Najman et al. 2017;Zhou et al. 2020). The equivalent Eocene successions are well exposed in the Northeast, India and represented by the Geku and Dalbuing formations of the Yingkiong Group (Arunachal Pradesh); Tura, Siju, and Rewark formations (Garo Hills), Shella and Kopili formations of the Jaintia Group (Jaintia Hills), Disang and Barail groups of Manipur, Nagaland and Assam.
The North Cachar Hills are part of southern Assambounded Nagaland, Mizoram, Manipur, and Tripura states. These hills are bordered on the west by the Kopili River, which is an important tributary of the Brahmaputra River and constitutes the eastern flank of the Shillong Plateau. These hills form the principal watershed between the Brahmaputra and Surma valleys. The Assam state consists of three structural units including the structural core of the region which is the Karbi Anglong plateau, and the Tertiary depositional zone and the alluvium depositional plains of Brahmaputra and Barak valleys. The pioneering work documenting the geology of Assam was done by Medlicott (1865), Mallet (1876), La Touche (1885), and Samanta (1968Samanta ( , 1970. Subsequently, a detailed account of the stratigraphy and structure of the Tertiary succession of Assam was done by Evans (1932) and GSI (2009). Wilson and Metre (1953) worked out foraminiferal biostratigraphy of the late Cretaceous-Eocene sequence of Assam. Samanta (1971) proposed a foraminiferal zonation of the early Tertiary sediments in and around Garampani area. Since 1972-73, geologists of the State Geology and Mining department, Assam carried out systematic geological mapping for locations of promising limestone deposits. The efforts of the Geology and Mining department, Assam resulted in locating one of the largest limestone deposits of the region in 1978.
This paper focuses on the detailed petrography and stable isotopes results, for the first time, of the Garampani limestone. The present investigation was conducted on the Middle Eocene succession for their classification, diagenetic, depositional environments, and climatic conditions.

Location and geological setting
The study area named Umrangso, (previously referred to as Garampani), is a part of Dima Hasao District, earlier called North Cachar Hill District, and located in the southern part of the state of Assam, in northeast India. In the study area, early Tertiary rocks (Eocene) rest unconformably over the weathered basement of Pre-Cambrian rocks. The Eocene sediments developed in and around study area are made up of three formations in ascending order, Mikir Formation, Garampani Limestone Formation, and Kopili Formation ( Fig. 1a; Table 1). The sequence is apparently conformable (Samanta 1971;Saxena and Trivedi 2009). The Garampani limestone and Kopili formation are highly fossiliferous and contain foraminifera. The studied limestone belongs to the Garampani Limestone Formation which is equivalent to the Sylhet Limestone Member of Shella Formation of Jaintia Group. The formation attains a maximum thickness of 120 m and is chiefly made up of fossiliferous limestone with a few layers of marl and shale. At a few localities, the Garampani limestone was found to rest directly on the Precambrian complex. Elsewhere it rests conformably on Mikir Formation, and is overlain by Kopili Formation (Samanta 1971). In the study area, a single limestone horizon underlain by sandstone represent the base and is considered as NE continuation of Nonghlie-Sutngali limestone belt of Meghalaya (Baishya 1994).
During the Eocene period, a thick pile of sediments was deposited over the Precambrian basement on shelf areas, in Assam. Intense tectonic activity in the Himalayan and adjoining Arakan Yoma region has resulted in Assam Shelf featuring numerous faults (Murty1983). Continuous activity along the fault zones and collision of Indo-Asian plate during late Palaeocene-early Eocene controlled the depositional regime in this region (Jauhri and Agarwal 2001;Tewari et al. 2010). Platform conditions were established during late Cretaceous period with onset of sandstone deposition. The clastic sedimentation carried on till the end of the early Paleocene, but a late Paleocene modification in the tectonic regime lowered the subsidence rates and led to a major reduction in the supply of clastic material. This event resulted in development of the carbonate platform upon which the carbonate sequences were deposited. The carbonate platform existed until the late middle/late Eocene when another alternation in the tectonic regime resulted in the increase in terrigenous clastic material and initiated deposition of thicker clastic sequences during the late Palaeogene and Neogene times (Jauhri and Agarwal 2001). The Jaintia to the east probably remained as landmass till early Eocene period and experienced subsidence only during the deposition of the coal bearing sediments. Then sedimentation continued uninterrupted over the submerged portion of the North Cachar Hills. Thus, the North Cachar Hills experienced marine transgression and regression with variable time lags. This has resulted in the formation of limestone horizon during the early Eocene period. During the late Eocene period, the area became a milieu of deeper marine environment when the thick unit of Kopili Formation accumulated (Sarkar 2015).

Materials and methods
The studied samples were collected during fieldwork from Jamunagar limestone quarry of Umrangso area (92°47′30.4E′′:25°30′58.8′′N). A detailed section of the carbonate succession was measured taking into account important features, such as the biogenic assemblages, lithology and sedimentology. A total of twenty (20) samples were collected from the section of approximately 70 m height. Thin sections were prepared using conventional method of thin section preparation for petrographic and diagenetic studies. Folk (1959Folk ( , 1962 and Dunham (1962) methods were followed for the classification of limestone. In the present study, sample no. JS-13 (shaly limestone) was analyzed palynologically based on standard techniques and identification has been attempted using earlier papers (Funkhouser and Evitt 1959;Staplin et al. 1960 (Friedman and O'Neil 1977). The carbon and oxygen isotopic data of representative samples are presented in Table 2.

Results
A total of nineteen (19) thin sections were studied for petrography and diagenesis, and are discussed under the following sub-headings.

Petrography
The Garampani Limestone is hard and compact, light gray to bluish gray in color with shale partings at places. Toward surface, these are found to be more argillaceous and at places ferruginous turning brownish gray to red-colored ( Fig. 1b: b). These limestones are highly fossiliferous with both macro-and micro-structures containing abundant foraminifer, predominantly Nummulites (( Fig. 1b: f). The environment of deposition is identified based on the framework constituents of limestone, which differ with physical energy.
Non-carbonate minerals, such as glauconite and iron, are present in some of the sample. Excluding the non-carbonate minerals, the framework constituents of the limestones are grouped as follows:

Allochemical constituents
The allochemical constituents in the studied limestone are skeletal grains, pellet, and intraclast. Fossils volumetrically occupy major portion ranging from 19.1 to 82.2% and are found mostly embedded in the groundmass of microcrystalline calcite matrix. The identified foraminiferal fossils are Nummulites, Assilina, Alveolina, and Discocyclina with subordinate coralline red algae and coral fragments (Fig. 2). These foraminifera are common larger benthic foraminifera (LBF) in Paleocene-Eocene Tethyan shallow-water carbonate platforms (Less1987), which are used extensively as a tool for reconstruction of paleoenvironment and paleo-bathymetric zones (Samanta 1971;Hallock and Glenn 1986;Braga and Aguirre 2004). These foraminifera and calcareous algae are typically associated with shallow-water carbonate sediments and are similar with those identified by earlier workers from different localities of Assam and Meghalaya (Gogoi et al. 2009;Sarkar 2017).
Volumetrically, pellets vary from 0.3 to 2.1%. In thin sections, they show remarkable uniformity of shape and size ( Fig. 2 g, h). Pellets are the polygenetic group of grains, typical of shallow, low-energy, warm supersaturated seas of restricted circulation at the time deposition (Tucker et al. 2005;Scoffin 1987). Intraclasts are found in negligible amount (0.8 to 1.1%) in the studied limestones. Though the amount is negligible, the presence of intraclast indicates gentle disturbance in the depositional environment and also it can be an evidence of shallow-water regime, as the bottom would have to be shallow enough to be periodically affected by the turbulence that causes intraclast formation because even intense storms have little effect below wave base. Intraclasts are regarded as the most important of all the allochem particles (Folk 1962) because of their implication of shallow water, lowered wave base or possible tectonic upliftment.

Orthochemical constituents
In the limestone, orthochems are represented by micrite and sparry calcite. Micrite is the dominant constituent and it varies from 69.70 to 99.12% and is believed to be deposited as biochemical ooze or produced by attrition of organic shells. However, some of the micrite may form in place as a diagenetic segregation or concretion (Folk 1962).
Sparry calcite forms crystals generally over 10 microns in diameter and is distinguished from the micrite by its clarity as well as coarser crystal size which may go up to 1 mm or more in diameter (Sander 1951). In the studied limestone, the percentage of sparry calcite ranges from 0.88 to 21.5 %.

Classification
A systematic classification of the limestone has been documented to decipher the paleo-environment and energy condition during the deposition of limestone. Folk (1959Folk ( , 1962 and Dunham (1962) carbonate classification were followed for purpose of environmental interpretation. Folk (1959) places primary emphasis in classification of the bulk carbonate rocks on the relative proportion of allochems, micrite, and sparry calcite cement and utilizes these three as end member of the triangular classification of limestone. Plotting the volumetric proportion of the framework constituents to the triangular diagram, the studied limestone belongs to Type I (Sparry Allochemical Rocks) and Type II (Microcrystalline Allochemical Rocks) as shown in Fig. 3. According to Folk (1962), Type I limestone represents deposition in strong persistent current in high-energy environment while Type II limestone indicates weak, short-lived currents and poorly winnowed sediment. He further subdivided these three major families into eight groups reflecting the textural spectrum to understand the energy condition of the depositional environment (Folk 1962). Two types of limestone have been identified viz. sparse biomicrite and packed biomicrite from the studied limestone (Table 3). Dunham's (1962) classification is based on the depositional fabric of carbonate rocks and the fundamental criterion for the subdivision is the nature of the framework grain of the sediment. He recognized four textural classes of limestone where each class signifies a relatively clear energy level during the deposition. Comparing the percentage of limestone constituents with Dunham's (1962) classification, three types of limestone have been identified as wackestone, packstone, and grainstone (Table 3). Wackestone is a mud-supported carbonate rock containing more than 10 percent allochems in a carbonate mud matrix. This type of limestone is common throughout the succession, although this type of limestone is more abundance in the lower and middle succession when compared with grainstone, indicating relatively lower energy condition at the lower and middle successions. Packstones are grains supported by muddy carbonate rocks where micrite is precipitated within the intergranular pores. This is the most common type in the studied limestone. Generally, grain supported rocks are deposited in agitated water while muddiness is generally a property of rocks deposited in quiet water. Exhibiting both properties in this type of rock might be early or late infiltering of previously deposited mud-free sediment or prolific production of grains in calm water (Dunham 1962). Most muddy limestones have undergone compaction caused by dewatering. In this way, some sediments deposited as wackestone can be converted to packstone (Shinn and Robbin 1983). Grainstone is the mud-free carbonate rock where grains are abundant and support one another. It is more dominant in the upper part of the succession. Non-carbonate mineral gluconite occurs as sub-rounded to rounded grain in the limestone. Its occurrence is seen to be restricted only within the small bands of limestone in the middle part of the succession. The glauconite may originate from clayey substances incorporated during cementation catalyzed by the enhancement of iron-bearing solution from the overlying formation and pressure solution.

Diagenetic aspects
The petrographic study of the limestone indicates that the visible diagenetic features that occurred under hydraulic conditions prevailed during burial include compaction, pressure solution, neomorphism, and recrystallization. Microstylolites and calcite veins showing two sets of cleavage are also very common and sometimes deposition of opaque minerals of unknown composition is observed. Compaction generally divided into two regimes, viz., mechanical and chemical compaction. However, these two effects are difficult to separate. The petrographic study allows some quantification of the respective roles of the mechanical and chemical compactions in natural carbonate through observation of microstructures (Meyers 1980;Gratier et al. 1999;Budd 2002). During the loading phase, chemical compaction occurs by pressure solution creep which is enhanced by the presence of cracks at the grain-to-grain contacts. The chemical compaction involves early meteoric and marine diagenesis as well as crack propagation in the presence of reactive fluid and dissolution-precipitation resulting from pressure solution. Two types of chemical compaction features are observed in the studied rocks that are solution seams at grain contact and stylolites (Fig. 2 m, n).
Pressure solution is an important process of porosity elimination in carbonates (Rutter 1983;Tada and Siever 1989). The role of pressure solution in the post-depositional alteration of carbonate sediments is that the CaCO 3 released by stylolite growth becomes major source of late diagenetic cement. The zig-zag form of the surface is presumed to be the consequence of lateral variations, along the interfaces of solubility difference of rocks. When a more soluble part confronts a less soluble part across a solution film, the more soluble part gets dissolved and the less soluble part moves into the resultant space. In the present case, the more soluble carbonate sediments were dissolved and moved out, whereas the insoluble residual non-mobile non-carbonate material got concentrated into the resultant space (Devi et al. 2016). Due to pressure solution effect, microstylolites are formed (Fig. 2n). Stylolites of simple and complex type are also observed in the studied rocks.
Neomorphism is an inclusive term to define, which is compositional and has replacive processes, such as recrystallization and inversion (Folk 1965). Neomorphic processes can play an active role in modifying the textural characters of carbonate sediments during both early and late diagenesis (Mukherji and Young 1973). In the studied limestone, aggrading neomorphism (crystal enlargement) is common (Fig. 2), leading to a coarser mosaic of crystals (microspar) as a product of neomorphism (Folk 1965). Microspar is necessarily a product of secondary recrystallization of a previously lithified micrite (Munnecke et al. 1997). It is clear that microspar, showing mosaic texture, must be of diagenetic origin.

Carbon and oxygen isotopes
The Garampani Limestone has slight variation in oxygen and carbon isotopes values ranging from −6.25 to −4.67 ‰ with an average value of −5.40‰ for δ 18 O (PDB) and from 1.64 to 0.76 0 / 00 (PDB) with an average of 1.23‰ for δ 13 C ‰ ranges respectively (Table 3). Paleo-temperatures are calculated using the following equation proposed by Craig (1965) O) 2 in which T, δ c 18 O, and δ w 18 O represent mean paleo-temperature, oxygen isotopic composition of the sample (V-PDB), and oxygen isotopic composition of sea water (SMOW) at the time of limestone precipitation. During Middle Eocene of this region, the climate generally was warmer compared with the overall cooling trend for Cenozoic era (Zachos et al. 2001). Depositional settings of the limestone can be estimated using the following equation proposed by Keith and Weber (1964), Z = a(δ c 13 C + 50) + b(δ w 18 O + 50) where 'a' and 'b' are 2.048 and 0.498. δ c 13 C is the carbon isotopic composition of the sample (V-PDB), and δ w 18 O the oxygen isotopic composition of the sample (V-PDB). The limestones with Z values above 120 are considered as marine, whereas Z values below 120 would be classified as freshwater type.
Thus, in the present study, all the samples observe Z values above 120 ranging from 126.00 to 128.05 presenting their marine origin (Table 2). This has been supported by the occurrence of LBF, dinoflagellate cysts, and microforaminiferal lining (Fig. 4). The bivariate plot between δ 13 C‰(PDB) and δ 18 O ‰(SMOW) of all samples is similar to marine limestone (Fig. 5). Furthermore, the bivariate plot of δ 13 C-δ 18 O with generalized isotopic fields for carbonate components, sediments, limestones, cements, dolomites, and concretions was established by Hudson (1977) and later modified by Nelson and Smith (1996). This bivariate plot (Fig. 6) of the Garampani Limestone falls in the field of marine limestone and margin of burial cement.

Discussion
As a whole, the Garampani limestone unit is highly fossiliferous with a predominance of foraminifera and algae as allochems. High frequencies of larger foraminifera are clearly visible as framework grains, which is similar to Jaintia Hills of Meghalaya (Matsumaru and Sarma 2010). The assemblage is characterized by relatively abundant Assilina species along with Nummulites and few Alveolina at the lower part of the succession, while the upper part is characterized by the presence of Discocyclina species and large, highly evolved Nummulites. On the basis of these larger foraminiferal assemblages, the studied limestone is dated as Middle Eocene. However, the abundance of genus Assilina in the lower and the absence in the uppermost part of the succession suggest the upper age limit of Garampani limestone is Middle Eocene age based on the last occurrence of the Paleocene-Middle Eocene genus Assilina (Samanta 1971). The shaly limestone found in the middle part of the succession contains well-preserved palynological assemblage of dinoflagellate cysts and foraminiferal linings (Fig. 4). The dinocysts consist of Glaphyrocystaintricate, Operculodinium centrocarpum, Impagidinium sp., and Spiniferites pseudofurcatus. Among these dinocysts, Glaphyrocystaintricate and Operculodinium centrocarpum are closely comparable with Middle Eocene palynological assemblage recorded in the Middle Eocene sediment throughout the globe (Sexana and Sarkar 2000;Singh et al. 2022). showing various isotopic fields (after Hudson 1977;Nelson and Smith, 1966) of a selection of carbonate components A petrographic property of limestone reveals the energy condition active in the basin at the time of deposition of sediments. Sparsed and packed biomicrite of Type I and Type II (Folk 1959(Folk , 1962 limestones having texturally wackestone, packstone, and grainstone nature show deposition of the studied limestone under strong or persistent to weak, shortlived energy condition. The Garampani Limestone shows the dominance of wackestone, and packstone toward the lower part and those grades to grainstone toward the topmost part. Dominant of micrite at the lower part normally correlated with shallow-water low-energy condition and grains at the upper, high energy at the time of deposition (Scoffin 1987). The phenomenon of grain supported in the upper part of the succession has also added importance of bearing diagenetic history of the sediment. The low percentage of sparite may have resulted from recrystallization of micrite during the neomorphism as aggrading neomorphism (Folk 1959). The microstylolite structures seen in the thin sections suggest that the pressure solution was caused during the early stages of diagenesis of the limestone as these structures do not cut across the fossil allochem grains. No evidence of resedimentation e.g., turbidites, related to a steep slope, dominance of micrite and absence of reef constructing organisms except few coral fragments in the entire succession suggests a low-energy setting environment, which is more likely to occur on a low-gradient slope characteristic of a ramp rather than a shelf (Burchette and Wright 1992;Wright and Bur-chette1998). The presence of glauconite in the rock suggests deposition in shallow marine condition of normal salinity with slightly reducing environment facilitated by organic matter and formed in areas of slow sedimentation (Velde 2003). Pellet facies associated with foraminifera especially Miliolina in the limestone suggest protected lagoon environment (Brigaud et al. 2009). Deposition in ramp environment is also supported by the presence of the Nummulites and the scarcity of alveolinids and orbitolitids in the matrix, indicative of deposition well below the fair-weather wave base (FWWB). The elongated Assilina and Discocyclina indicate the relatively deep water on more open parts of the ramp (Racey1994).
The Middle Eocene of the study area had considerably higher sea surface temperature (SST) (average 37.48 °C) in the past (Shackleton 1986). The Miocene was also significantly warmer with the continuation of Himalayan orogeny and initiated during early Eocene (38Myr), vast expansion of the west Antarctic ice sheet and enhanced monsoonal circulation (James et al. 1994;Retallack 2001). The carbon and oxygen isotopes composition did not undergo major changes during diagenesis. The stable isotope values of this region have two distinct trends. The δ 18 O (PDB) values remain negative through this Eocene section and have more or less the same values except for a slight enrichment on sample nos., JS-12 and JS-5 (Table 2; Fig. 7) due to warm conditions. The δ 13 C values also remain positive with slight difference values in the (Eocene section). δ 13 C values of the samples have both increasing (sample nos. JS-1, JS-2, JS-3) and decreasing trends (middle and top parts of section represented by  with an irregular pattern of excursions in other parts of section (Table 2; Fig. 7). During transgression, more organic matter is trapped in the marginal areas, resulting in the enrichment of δ 13 C values, whereas during regression, the trapped organic matter is eroded and oxidized, resulting in depletion of δ 13 C in the deep ocean (Fig. 7). Thus, the high algal population and photosynthetic activity in the depositional environment in a shallow marine margin suggest positive δ 13 C values (Milliman and Muller 1977;Nelson 1988). All samples are represented by positive values of δ 13 C. The irregular patterns of positive δ 13 C values can be correlated with the transgressive phase during the deposition of the Garampani Limestone and also indicated that the limestone samples were not subjected to a major degree of diagenesis (Narayanan et al. 2007). The presence of agal in the shale-rich limestone sample (JS-13) identified under palynological slide might be responsible for the positive δ 13 C values. Diagenesis often results in more negative δ 18 O values in marine carbonate due to cementation and recrystallization (Land 1970;Allan and Mathews 1977). The high temperature indicated by δ 18 O values suggests that the Middle Eocene was generally marked by warmer climatic condition.

Conclusions
Based on petrographic characteristics, diagenetic signatures and isotopic significances, the following conclusions are summarized: 1. Petrographic constituent reflects active energy condition during deposition. Sparse and packed biomicrite limestone of texturally wackestone, packstone, and grainstone nature indicates the environment of deposition was a short-lived period of decreasing energy. Presence of non-carbonate constituent in the rock indicates shallow marine with slow sedimentation condition, which is a characteristic of ramp environment. 2. Diagenesis played a major role in modifying the original textural and compositional features of the investigated limestone due to marine burial characterized by compaction, recrystallization, pressure solution, neomorphism, etc., which affect the rocks in different diagenetic environments. Even though the carbon and oxygen isotopes composition did not undergo major changes during diagenesis, however, negative δ 18 O(PDB) values (− 6.25 to −4.67) suggest result of marine diagenesis.
3. Occurrence of stratigraphically significant foraminiferal assemblages, such as Assilina, Nummulites Discocyclina and palynofossils, particularly Glaphyrocystaintricate and Operculodinium centrocarpum assigned the age of Garampani limestone as Middle Eocene. Elongated species of Assilina and Discocyclina indicate the relatively deep water on more open parts of the sea. Presence of nummulite and scarcity of alveolinids and orbitolitids among the foraminiferal assemblages indicates deposition takes place below the fair-weather wave base (middle ramp). Pellet facies associated with foraminifera especially Miliolina in the limestone suggest protected lagoon environment (inner ramp. In addition to this, absence of reef constructing organism in the entire succession indicates the sediments were deposited in shallow marine ramp environment which is type of carbonate platform.