This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/authorsrights Author's personal copy Available online at www.sciencedirect.com ScienceDirect Palaeoworld 22 (2013) 119–132 The Miocene coastal vegetation of southwestern India and its climatic significance Andrea K. Kern a,∗ , Mathias Harzhauser b , Markus Reuter c , Andreas Kroh b , Werner E. Piller c a State Museum of Natural History Stuttgart, Rosenstein 1, 70191 Stuttgart, Germany b Natural History Museum Vienna, Burgring 7, 1010 Vienna, Austria c Institute of Earth Sciences – Geology and Palaeontology, Graz University, Heinrichstrasse 26, 8010 Graz, Austria Received 6 May 2013; received in revised form 14 September 2013; accepted 4 October 2013 Available online 18 October 2013 Abstract The late early to middle Miocene sediments at Varkala cliff section (SW India) offer a great potential to study paleoenvironmental and paleoclimatic conditions. The succession consists of coastal marine sands alternating with fine grained, organic-rich siliciclastics. Based on several palyno-samples, a mangrove flora could be reconstructed, composed of Rhizophoraceae, Avicennia, Xylocarpus, and Sonneratia. These grew in front of backswamps, followed by evergreen forests in hinterland with many rainforest elements. Despite these similarities between the Miocene and the modern vegetation, differences can be documented, e.g., the abundance of Nypa and the absence of gymnosperm in the Miocene vegetation. Based on the Coexistence Approach, a temperature is reconstructed, which is similar to today’s (mean annual temperature 22.2–27.5 ◦ C; coldest month temperature 20.6–23.6 ◦ C; warmest month temperature 27.5–28.1 ◦ C). However, in comparison to today, significantly lower values for the mean annual precipitation are reconstructed (1748–1958 mm), although a seasonality pattern was established (wettest month mean precipitation 225–358 mm; driest month mean precipitation 18–83 mm). Comparison with other localities along the west coast of southern India hints at a similar vegetation and climate throughout the early and middle Miocene. Furthermore, the same results were obtained for SE India. This is in contrast to the strong west-east gradient in rainfall caused by the Indian Monsoon nowadays. Therefore, the paleovegetation of the early and middle Miocene of southern India appears to be more uniform than today. This outlines the complexity of describing the paleoclimatic evolution in the tropics based on plant fossils, but emphasizes the need for further study of this area to understand the early history of the Indian Monsoon. © 2013 Elsevier B.V. and Nanjing Institute of Geology and Palaeontology, CAS. All rights reserved. Keywords: Indian Monsoon; South Asian Monsoon; Paleoclimate; Paleovegetation; Miocene 1. Introduction Southwestern India is known today as one of the biological hotspots along the mountain chain of the Western Ghats (Myers et al., 2000). The recent high biodiversity is boosted by the fact that the Western Ghats acted as a retreat area for many plants during phases of climatic change such as the ice ages (Farooqui et al., 2010). The history of this tropical area can be reconstructed based on the excellently preserved fossil floras in ∗ Corresponding author. Tel.: +49 7118936168. E-mail addresses: andrea.kern@smns-bw.de (A.K. Kern), mathias.harzhauser@nhm-wien.ac.at (M. Harzhauser), markus.reuter@uni-graz.at (M. Reuter), andreas.kroh@nhm-wien.ac.at (A. Kroh), werner.piller@uni-graz.at (W.E. Piller). the district of Kerala. Therefore, many previous studies were conducted on these sediments which focus on the paleovegetation from Paleocene to Miocene times (e.g., Venkatachala and Rawat, 1972, 1973; Rao and Ramanujam, 1975; Ramanujam, 1982; Rao, 1995). These investigations reveal that tropical vegetation similar to the modern one did exist since the Paleocene in SW India. However, dating of many floras remains difficult and the similarity of the fossil vegetation often does not provide a good indication of the age of the locality. One exception is a 21m-long cliff section near the village Varkala (Fig. 1), which was recently correlated to the early to middle Miocene in age (Reuter et al., 2013). Herein, we discuss the depositional environment and paleoflora of the Varkala cliff section in order to reconstruct the paleoenvironment as well as paleoclimate using the Coexistence Approach (CA, Mosburgger and Utescher, 1997). These results are further compared with the recent vegetation and 1871-174X/$ – see front matter © 2013 Elsevier B.V. and Nanjing Institute of Geology and Palaeontology, CAS. All rights reserved. http://dx.doi.org/10.1016/j.palwor.2013.10.001 Author's personal copy 120 A.K. Kern et al. / Palaeoworld 22 (2013) 119–132 Formation (Raha, 1996; Vaidyanadhan and Ramakrishnan, 2008). The latter indicates seagrass environments during a marine transgression into the marginal marine lagoons and swamps (Reuter et al., 2011). Calcareous nannoplankton from the Padappakkara type locality belongs to nannoplankton zone NN3 (middle Burdigalian) for the Quilon Formation (Reuter et al., 2011). Therefore, a late Burdigalian to middle Miocene age of the Ampalapuzha Formation can be assumed (Reuter et al., 2011, 2013). 2.2. Lithofacies of the study locality Fig. 1. Map of the southern Indian Peninsula showing the geographical position of all localities mentioned in the text (underlined names next to white dots). Varkala cliff section marked with a white star. modern climatic conditions in Kerala, as well as various paleofloras from several early to middle Miocene localities along the coastline of Kerala in western India (Rao and Ramanujam, 1975; Ramanujam, 1982; Rao, 1995) and from coeval sites in southeastern India (Singh et al., 1992; Mukherjee, 2012) to reveal differences or similarities (Fig. 1). Although the overall composition of the paleovegetation is already well known from this area, modern analyses of the paleoclimate are scarce (Reuter et al., 2013). Such studies could be of particular interest to answering the questions about the origin and history of the Indian Monsoon. In particular, there are only a few papers on the manifestation of an early Indian Monsoon (Clift et al., 2008; Srivastava et al., 2012; Reuter et al., 2013) in contrast to many new studies about the evolution of the East Asian Monsoon in the Paleogene/early Neogene (e.g., Guo et al., 2002; Sun and Wang, 2005; Wan et al., 2007; Quan et al., 2012). 2. Geology 2.1. Geological background and stratigraphy The studied outcrop is located in the onshore Kerala Basin at the coastal cliffs near the village Varkala in the Kerala district, SW India (Fig. 1), 420 m SE of the “Hindustan Beach Retreat” hotel (N 08◦ 43′ 47′′ , E 76◦ 42′ 30′′ ) and exposes a 21 m section along the coastline. These sediments belong to the Ambalapuzha Formation (Raha, 1996; Vaidyanadhan and Ramakrishnan, 2008; Reuter et al., 2013; Fig. 2) of the Cenozoic Warkalli Group, which is characterized by siliciclastics with interbedded lignite seams (Campanile et al., 2008; Vaidyanadhan and Ramakrishnan, 2008). According to Rao and Ramanujam (1975) and Rao (1995), palynofloras and ostracod assemblages indicate marginal marine, brackish lagoons and brackish and freshwater swamps, although neither provides indications about the age of the sediments. However, the Ambalapuzha Formation conformably overlies the siliciclastic carbonatic Quilon Sediments in the Varkala cliff section range from quartz sands to clayey silts and silty clays (Fig. 2). Black organic-rich clays with variable contents of silt and poorly-sorted sands predominate (beds 1, 3, 4, 6, 8, 11, 12, 14, 17, top 23, top 24). This facies comprises dense standings of < 20 cm long and < 0.5 cm wide, partly iron oxide-coated, vertical tubes (beds 3, 4, 8, 11, 12, 14) (Fig. 3F). They represent rootlets as indicated by unidentifiable lignite preserved in some of these tubes. Other beds are formed by laminated and interlaminated silty clays and clayey silts with subsidiary laminae of well-sorted fine-grained quartz sand and sand-lensed mud (beds 10, 20) or ripple-laminated quartz sands and silty sands with silty clay and clayey silt drapes and some lamination and flaser-bedding (beds 19, 21, 22) (Reuter et al., 2013). The sand deposits of beds 13 and 18 exhibit plane, subhorizontal laminae. Individual laminae are < 1 cm thick and consist of well-sorted coarse sand or fine gravel. Intercalated in bed 13 is a distinct 20 cm thick red horizon. It exhibits elongated protuberances at its irregular base, which extend a few centimeters into the underlying sandstone and are filled with red sediment. The thickness of bed 18 decreases abruptly from 1.2 m to 55 cm at a sharp surface with a steep sub-vertical dip to the west, which is filled by the sediments of layer 19 on top (Fig. 2). Bioturbation occurs in all facies (Fig. 3C, D). Dominant ones are up to 25 cm deep vertical burrows with the circular cross-sections (5 cm). Well-preserved casts of these burrows (Diplocraterion ichnofossils) exhibit chevron patterns and ridges on their surface (Reuter et al., 2013). Lignitic woods (up to 30 cm length) also occur across the entire section but are most common in the fine-grained facies (Fig. 3B). Amber pieces of up to 6 cm long were scattered in the sand of bed 7 as well as at the top of bed 23 (Fig. 3E), where they are associated with small lignite fragments in black organic-rich, sandy silty clays at the top of bed 23 (Fig. 3E). The top of Varkala section is a 7-m-thick laterite bed (Fig. 3A). 3. Material and methods All eight samples were taken from dark organic rich layers and were processed at the Geological Survey of Austria in Vienna (labeled VC9-1, VC9-3, VC9-6, VC9-8, VC9-10, VC912, VC9-17 and VC9-20, referred later to as sample 1, 3, 6, 8, 10, 12, 17 and 20; Fig. 2). After the removal of the sand, the preparation followed the procedure of Klaus (1987) and Erdtman (1954). Finally, samples were sieved with a 6 ␮m nylon Author's personal copy A.K. Kern et al. / Palaeoworld 22 (2013) 119–132 121 Fig. 2. Lithological log of the Varkala cliff section with the position of the samples. sieve (also see Reuter et al., 2013). All materials are kept in the collection of Natural History Museum, Vienna. Each sample contained pollen grains; at least 200 were identified based on publications of fossil pollen (e.g., Saxena, 1982; Ramanujam, 1987; Singh et al., 1992; Rao, 1995; Misra et al., 1996; Mukherjee, 2012) and modern taxa (e.g., Rao and Lee, 1970; Thanikaimoni, 1987). Results are illustrated in a pollen diagram using Grimm (2004) (Fig. 4). Paleoclimate was estimated by the nearest living relative (NLR) based on method of the CA (Mosburgger and Utescher, 1997). For each taxon, its most appropriate modern reference taxon and its recent climatic distribution range are determined. Climatic parameters for mean annual temperature (MAT), the warmest month temperature (WMT) and the coldest month temperature (CMT), the mean annual precipitation (MAP) and for the mean monthly precipitation of the driest month (MPdry), the wettest month (MPwet) and the warmest month (MPwarm) were extracted from the Palaeoflora data base (www.palaeoflora.de) and compared between the taxa and different climatic parameters by using the software ClimStat (Mosburgger and Utescher, 1997). For each climatic value, intervals were generated, where all/a majority of the taxa coexist (coexistence interval). 4. Results 4.1. Pollen Most of the previous studies on the palynoflora of the Kerala sections applied parataxonomy and provided names based on pollen morphology (e.g., Rao and Ramanujam, 1975; Author's personal copy 122 A.K. Kern et al. / Palaeoworld 22 (2013) 119–132 Fig. 3. (A) Picture of the Varkala cliff section. (B) Large wood fragments, often containing amber. (C, D) Bioturbation. (E) Large piece of amber. (F) Sandy-silty clay with rootlets (presumably Avicenniaceae). Fig. 4. Pollen diagram drawn with the Tilia/Tiliagraph-program (Grimm, 2004). The number of taxa limited to the ones exceeding 2%, including spores. Values of mangroves, mangrove-associated plants, and aquatic taxa displayed according to the groups in Table 1. Author's personal copy A.K. Kern et al. / Palaeoworld 22 (2013) 119–132 Ramanujam, 1987; Rao, 1995). Nevertheless, the authors suggested modern botanical affinities for the majority of taxa, which were the basis for the comparison with other fossil and modern literature (e.g., Thanikaimoni, 1987). A summary of the palynological samples from the Varkala cliff section is already published in Reuter et al. (2013), listing 10 genera and 39 families (with a total of 49 taxa, Table 1). Spores of fungi and the rare remains of dinoflagellate cysts were not considered in the analysis. Spores were common in all samples ranging between 8.7% and 13.3% of the pollenspore assemblages (Table 1; Fig. 4). The pollen diversity in each sample ranges from 32 to 39 taxa. The composition of the assemblages varies only slightly between each sample. A total lack of gymnosperm pollen grains in all investigated samples agrees with the previous studies of this area (e.g., Rao and Ramanujam, 1975; Ramanujam, 1987; Rao, 1995). Arboreal angiosperms dominate over herbs with Arecaceae as the most diverse family. Arecaceae pollen are also the most abundant, reaching to a peak of 38.8% in sample 3 (Table 1; Fig. 4). Among them, the mangrove-associated taxon Nypa (Fig. 5D) is the most frequent one, resulting in values between 1.5% and up to 12.8% of the total assemblage. Next to palms, true mangrove plants are most common: Rhizophoraceae pollen (Fig. 5E) are represented by 6.5–13.5%, Avicennia (Avicenniaceae) (Fig. 5A) by 1.1–7.2%, and Sonneratia (Sonneratiaceae) by 0.7% up to 6.5%. Furthermore, regular occurrences of Sapotaceae (3.4–12.9%) (Fig. 5F), Combretaceae (4.2–8.7%) (Fig. 5D), and Euphorbiaceae (3.3–5.7%) are recorded. Meliaceae (0.2–6.9%) (Fig. 5C), Myrsiaceae (1.7–4.6%), and Rutaceae (1.6–5.1%) are in low amount in each sample. The only constant trend observed along the Varkala cliff section is the strong increase of Rubiaceae to up to 9.7%. Additionally, the samples at the bottom of the sequence contain more Arecaceae (excluding Nypa) and less pollen of mangroves (especially Rhizophoraceae and Sonneratia) (Fig. 4). This is related to a slight alternation between mangrove dominated (6, 8, 10, 17, 20) and forest-vegetation (1, 3, 12) samples. All detected taxa are listed in Table 1, the most abundant ones are shown in Fig. 4, and the most significant ones are illustrated in Fig. 5. 4.2. Paleoclimate estimates 4.2.1. Paleoclimate of the Varkala cliff section A climatic reconstruction based on the CA of flora from the Varkala cliff section has already been published by Reuter et al. (2013) and is presented here in Table 2. However, updates of the climatic information stored at the palaeoflora data base (www.palaeoflora.de) resulted in slightly different coexistence intervals (Table 2 and Appendix A). Main differences between the CA results in Reuter et al. (2013) and the revised results discussed herein are caused by updated climatic data for the palm Calamus, which was limiting several coexistence intervals in the previous version (MAT maximum value, CMT maximum value, MAP maximum value, MPdry maximum value). No major changes occur in comparison with the climatic intervals presented here, although the new climatic values often give slightly broader coexistence intervals 123 (Table 2). In addition, next to the aquatic taxon Myriophyllum, the palm genus Oncosperma caused problems for several coexistence intervals. Although Oncosperma would improve the interval of the MAT and MPwarm, it is an outlier for CMT, WMT, MAP, and MPwarm in the analyses, including the palynoassemblage from all samples. In addition, during single sample analysis with a lower diversity, Oncosperma strongly shifts interval borders, especially concerning the MAP, and was therefore excluded from the CA of all samples. The reason is unclear why Oncosperma causes these problems in the CA. This genus is distributed today in the tropics of south and southeastern Asia (Partomihardjo et al., 1992; Johnson, 1998; Baker and Couvreur, 2012) with five species. A possible link could be the uncertain natural distribution due to the long history of human use as food and building material as mentioned for Oncosperma tigillarium (Partomihardjo et al., 1992) or the threatened habitats of Oncosperma platyphyllum of and Oncosperma fasciculatum (Johnson, 1998; Madulid, 1998), which may therefore result in a possible non-adequate documentation of its natural climatic range. New CA estimates suggest an MAT for the Varkala cliff section of 22.2–27.5 ◦ C, with a CMT of 20.6–23.6 ◦ C, and a WMT of 27.5–28.1 ◦ C (Table 2). Since Dipertocarpaceae are distributed in climates with the lowest WMT of 28.1 ◦ C, the data suggest that the top end of the interval may be more likely to represent paleoclimatic conditions. Especially since this uncertainty could be linked to the fact that tropical climate data and distribution maps are often scarce, it results in less solid entries in the climatic data base. The MAP ranges between 1748 mm and 1958 mm as defined by Brownlowioideae and Chenopodiaceae (Appendix B). A division into ambiguous coexistence intervals was generated for the mean precipitation of the wettest month (MPwet), in accordance with the requirements of the palm Nypa. Today it is distributed in south, southeastern, and eastern Asia (Badve and Sakurkar, 2003; Mehrotra et al., 2003; Ellison et al., 2010) and the northern part of Australia (Dove, 2010), although it is recorded globally during the Cenozoic. Excluding Nypa would result in an MPwet interval between 225 mm and 320 mm confronted by slightly wetter conditions otherwise (225–358 mm). Nevertheless, the combination of two ambiguous coexistence intervals to one larger interval is an accepted application in the CA. Nypa, however, is one of the few elements that does not occur in the recent vegetation of Kerala but shares the similar space with many of the detected plants in southeastern Asia (Badve and Sakurkar, 2003; Mehrotra et al., 2003) and Australia (Dove, 2010). There are many reasons why Nypa is not distributed anymore in India and why it causes some problems in the CA. Badve and Sakurkar (2003) pointed out that Nypa today is very sensitive to changes in salinity, which is commonly related directly to the incoming fresh water from precipitation. This may lead to problems for these plants during the long dry phase in SW India (Badve and Sakurkar, 2003). However, deviating climatic requirements may also be caused by an anthropogenically-induced recent distribution, which would lead to wrong climatic entries in the data base. Otherwise, the ecological climatic requirements of the genus Nypa might Author's personal copy 124 A.K. Kern et al. / Palaeoworld 22 (2013) 119–132 Table 1 Percentages of all recorded plant taxa according to vegetation types such as primary “mangroves”, “mangrove-associated plants”, “aquatic plants”, and “forest vegetation”. Varkala (Kerala, India) Mangroves (presumed/solely) Avicennia Xylocarpus Nypa Rhizophoraceae Sonneratia Mangrove-associated plants Acanthaceae Brownlowia Caesalpiniaceae Combretaceae Malvaceae Aquatic plants Myriophyllum Plumbaginaceae Potamogetonaceae Typhaceae Forest vegetation Agavaceae Anacardiaceae Apiaceae Apocynaceae Arecaceae Asteraceae Bombacaceae Calamus Chenopodiaceae Clusiaceae Ctenolophaceae Dipterocarpaceae Droseraceae Euphorbiaceae Fabaceae Gunneraceae Iridaceae Lamiaceae Loranthaceae Menispermaceae Metroxylon Moraceae Myrsinaceae Myrtaceae Olacaceae Oncosperma Poaceae Polygalaceae Proteaceae Rubiaceae Rutaceae Sapindaceae Sapotaceae Symplocos Thymelaeaceae Mangroves Mangrove-associated plants Aquatic plants Forest vegetation VC09-1 VC09-3 VC09-6 VC09-8 VC09-10 VC09-12 VC09-17 VC09-20 3.4 0.4 2.8 10.2 1.3 1.6 0.2 6.4 8.2 0.7 1.3 1.3 12.8 6.6 6.6 2.3 1.7 6.6 13.5 3.7 4.0 1.4 7.1 9.2 1.7 1.1 0.7 4.8 8.5 3.7 4.1 1.5 5.5 13.0 1.3 7.2 6.9 1.6 7.9 2.8 0.9 0.0 0.4 8.7 0.8 1.3 0.2 2.4 4.7 0.7 1.6 0.0 1.3 7.5 1.0 1.2 0.0 1.4 5.2 2.3 1.4 0.0 1.7 4.3 0.7 0.0 0.4 2.2 5.5 0.4 1.1 0.4 1.5 8.1 1.7 0.0 0.3 1.6 6.3 0.3 2.5 1.7 0.0 0.4 0.7 1.8 1.1 0.0 1.3 1.6 0.0 0.0 1.2 0.9 0.0 0.6 1.7 1.2 0.0 0.0 0.7 1.8 1.1 0.0 0.0 1.1 0.4 1.3 1.3 3.8 0.9 0.0 0.9 1.9 0.9 6.4 21.6 1.1 0.9 1.3 0.0 1.5 0.2 0.0 0.0 5.1 0.4 0.2 0.9 1.1 0.0 0.0 0.0 3.4 1.7 0.8 0.0 5.9 0.0 0.0 0.0 2.8 2.6 0.0 3.4 0.6 0.8 0.0 0.2 0.0 1.6 25.7 0.7 0.0 2.4 0.2 1.8 0.4 0.0 0.0 5.8 0.9 0.0 0.7 1.6 0.0 0.2 0.9 1.1 4.4 1.6 0.4 3.3 0.4 0.0 1.1 1.8 2.9 0.2 9.1 0.0 0.7 0.0 2.0 0.7 1.3 16.4 0.0 0.3 3.3 0.0 2.0 0.0 2.0 0.0 3.3 0.3 0.0 0.0 0.3 0.0 0.3 0.0 3.6 3.0 0.0 0.3 2.0 0.7 0.0 1.0 0.7 1.6 0.0 11.8 0.0 0.3 0.0 0.9 0.0 1.2 21.0 0.0 0.3 0.9 0.3 1.7 0.0 2.0 0.0 6.1 0.0 0.6 0.0 0.0 0.0 0.0 1.2 1.2 4.6 0.6 0.0 1.4 1.2 0.6 1.4 1.4 1.7 0.0 9.2 0.0 0.0 0.0 0.7 0.0 1.2 18.9 0.2 0.5 2.1 0.0 0.5 0.5 1.7 0.5 3.3 0.5 1.4 0.5 0.5 0.0 0.0 2.1 5.0 1.9 0.7 0.0 5.2 0.7 0.7 1.2 1.7 1.7 0.0 10.6 0.0 1.4 0.0 1.5 0.7 0.4 22.8 0.0 0.0 1.1 0.7 0.0 0.4 1.5 0.0 4.4 0.0 1.1 0.0 0.7 0.0 0.0 3.3 2.6 2.6 0.4 0.0 3.7 0.7 0.4 1.1 2.9 2.9 0.0 12.9 0.0 0.4 0.0 0.9 0.4 2.1 18.3 0.0 0.6 0.4 0.0 0.2 0.0 0.0 0.6 5.3 0.9 1.1 0.0 0.6 0.2 0.2 2.1 0.6 3.0 1.1 0.2 0.6 0.0 0.0 0.0 8.7 5.1 0.0 4.9 0.2 0.4 0.0 1.6 0.0 1.9 11.3 0.0 0.6 0.6 0.0 0.9 2.8 0.0 0.3 3.5 0.0 0.9 0.0 0.0 0.0 0.0 0.0 6.0 2.5 0.6 0.0 0.3 0.0 2.5 1.3 9.7 1.9 0.0 8.8 0.0 0.9 18.1 10.8 4.5 66.5 17.1 9.3 3.5 70.1 28.5 11.5 3.0 57.0 28.0 10.1 2.6 59.4 23.4 8.0 2.8 65.7 18.8 8.5 3.7 69.1 25.4 12.8 2.8 59.1 26.4 8.5 6.0 59.1 Author's personal copy A.K. Kern et al. / Palaeoworld 22 (2013) 119–132 125 Fig. 5. Photos of typical and significant pollen. (A) Avicenniaceae. (B) Combretaceae. (C) Meliaceae. (D) Nypa. (E) Rhizophoraceae (2 pollen grains). (F) Sapotaceae. have changed during evolution with a slightly different climatic requirement as the nearest living relative. The MPdry is reconstructed based on one single limiting taxon on each end (Brownlowioideae–Chenopodiaceae). As the lower end appears less problematic, the presence of Chenopodiaceae narrows the coexistence interval from 132 mm down to 83 mm. Thus, the reconstruction suggests a dry season with at least half of the precipitation of the wettest month(s) as these are characterized by rainfall ranging around 225–358 mm. This indicates seasonality in the annual precipitation, although the range remains vague due to the two wide coexistence intervals of the MPdry and MPwet. Besides, the MPwarm does not suggest a clear connection between temperature and rainfall within the year. The comparison of the eight different analyzed samples along the Varkala cliff section does not indicate a clear climatic trend. The samples with a lower diversity (1, 6, 8) have wider coexistence intervals, in particular for CMT and MPdry. Very large MAP intervals (up to more than 3000 mm) are the result of the absence of Chenopodiaceae and Brownlowia (Table 2; Fig. 6). Both of these occur in low percentages during the whole section (Table 1), as Chenopodiaceae may most likely have grown far away from the deposition side in less salty water and Brownlowia produces only a low number of pollen grains (Thanikaimoni, 1987). Therefore, the absence of these elements may presumably be linked to taphonomy instead of disappearance of these plants in the vegetation or a presumable climatic shift. The presence of Chenopodiaceae in some samples (8, 12) further indicates a shift towards a more expressed dry month (maximum precipitation of Fig. 6. Climatic estimations for each sample of the Varkala cliff section focusing on precipitation values based on the Coexistence Approach, showing the whole coexistence intervals and mean values (thick line). Abbreviations: MAP: mean annual precipitation; MPwet: mean precipitation of the wettest month; MPwarm: mean precipitation of the warmest month; MPdry: mean precipitation of the driest month. Author's personal copy 126 217.0 175.0 110.0 163.0 21.7 23.9 18 taxa 52 taxa 25.8 24.2 13.6 19.4 23.6 23.5 27.5 27.5 28.1 27.8 1215.0 1748.0 1958.0 1864.0 322.0 270.0 349.0 320.0 19.0 18.0 59.0 56.0 175.0 175.0 221.0 146.0 146.0 39.0 22.2 25.8 16.8 36 taxa 24 taxa 17 taxa 26.8 26.8 25.8 21.1 21.1 10.6 23.5 23.6 23.5 27.5 27.5 22.2 27.8 27.8 28.1 1748.0 1748.0 252.0 2099.0 1958.0 1958.0 270.0 270.0 84.0 320.0 320.0 358.0 18.0 18.0 2.0 56.0 56.0 59.0 175.0 175.0 114.0 114.0 22.2 22.2 West Varkala Varkala (Reuter et al., 2013) Quilon beds Warkalli beds Kalarakod-Nirkunnam East Panruti Neyveli lignite 29 taxa 27 taxa 27.5 26.6 20.6 20.6 23.6 22.8 27.5 28.1 28.1 28.1 1748.0 1748.0 1958.0 1864.0 225.0 225.0 358.0 358.0 18.0 18.0 83.0 55.0 MP-warm max MP-warm min MP-dry max MP-dry min MP-wet max MP-wet min MAP max MAP min WMT max WMT min CMT max CMT min MAT max MAT min Taxa analyzed Coexistence approach Table 2 All calculated climatic estimates for temperature, precipitation, and seasonality based on the CA for each of the localities (based on Rao and Ramanujam, 1975; Ramanujam, 1987; Singh et al., 1992; Rao, 1995; Mukherjee, 2012; Reuter et al., 2013). Additional information is available in the Appendix A. A.K. Kern et al. / Palaeoworld 22 (2013) 119–132 83 mm) (Fig. 6) and/or a lower maximum border of the interval for MPwet (358 mm) in samples 3 and 12 (Appendix A). All these shifts in the precipitation values are illustrated in Fig. 6; however, they seem to be in connection with the diversity of each sample and do not reflect real oscillations in the climate. 4.2.2. Climate of the SW coast (Kerala) To compare the local climatic data from Varkala cliff section, published plant lists were analyzed using the CA to observe possible regional differences. A summary of the occurring plants in different localities of the Warkalli Formation (Varkalai/Warkalli, Alleppy area and Cannanore beach) in the different regions of Kerala was presented in Ramanujam (1987) along with a list of the determination of the NLR of each morpho-palyno-taxon (Appendix B). The herein discussed Varkala cliff section most likely represents the mentioned Varkalai/Warkalli locality in Ramanujam (1987), which is the type locality of the Warkalli Formation. However, several outcrops occur along the coast of Varkala and the exact origin of the different samples is not clear. As mentioned by Ramanujam (1987), all localities of the Warkalli Formation show the same type of vegetation. The occurring plants are to a large extent comparable to the Varkala flora discussed in this paper and in Reuter et al. (2013), although a few additional plants are listed (Araliaceae, Coprosma, Eugenia, Ilex, Leea, Nyssa, Myricaceae, Vitaceae, Oleaceae) while others are missing (Malvaceae, Myrtaceae and Gunneraceae). This resulted in 24 taxa with NLR significant for the CA. In addition to the aquatic plants Potamogeton and Myriophyllum, Fagaropsis was excluded from the CA. Ramanujam (1987) suggested that Retitricolporites rhombicus may possibly be Fagaropsis; however, his observation was based only on light microscopy. Fargaropsis is not distributed in India, south or south-eastern Asia today, and it occurs only in tropical eastern Africa. There it is often limited to higher altitudes (e.g., Hitimana et al., 2004; Munishi, 2007; Hemp, 2011), which could explain why it acts as an outlier and suggests climatic values appearing too cold and too dry in the CA. However, Fargaropsis had distributed from Africa to Europe and Asia (Martinetto, 2001; Ling et al., 2009) during the Cenozoic and therefore could have been adapted to broader climatic conditions as well. Although fewer NLRs were included in the CA, the coexistence interval is narrower (Table 2), especially because of the occurrence of Coprosma and Leea (MAT, CMT, WMT, MPwet, MPdry, MPwarm). Coprosma (Rubiaceae) is a tropical and subtropical plant, and is not present in the southern Indian vegetation today (e.g., Barboni and Bonnefille, 2001), similar the topical liana Leea (Vitaceae). In comparison to the Varkala cliff section, the MAT interval is moved towards warmer values (25.8–26.8 ◦ C), whereas the CMT limits down to a colder month (21.1–23.6 ◦ C), as opposed to the warmest month of 27.5–27.8 ◦ C. The MAP values are supported by the presence of Chenopodiaceae and Coprosma, which clearly suggests less than 2000 mm (1748–1958 mm). The MPwet shifts to values lower than 320 mm, excluding the problematic taxon Nypa (as discussed above in 4.2.1.) while the MPdry signalizes an even drier month than in Varkala (18–56 mm). Furthermore, the rainy season Author's personal copy A.K. Kern et al. / Palaeoworld 22 (2013) 119–132 is shifted towards the warmer period of the year (MPwarm 146–175 mm) as indicated by Leea vs. Brownlowiodeae (Table 2 and Appendix A). A second list of three outcrops located close to each other in the south of Kerala (Padappakkara, Paravur and Edvai) was discussed in Rao and Ramanujam (1975). These belong to the Quilon Formation (middle Burdigalian), which is underlying the Warkalli Formation (Reuter et al., 2011). Here, the CA of 36 taxa resulted in these similar coexistence intervals as those in the Varkala cliff section and in Ramanujam (1987) (Table 2). Despite the age difference, vegetation, and likewise climate, is highly comparable. While the temperature reconstructions vary only within one degree Celsius in comparison to Varkala (MAT: 22.2–26.8 ◦ C; CMT: 21.1–23.5 ◦ C; WMT: 27.5–28.3 ◦ C), the seasonality pattern in rainfall is more intensively presented as in the list of the Wakalli beds localities. Also with this paleofloara a significantly drier month could be reconstructed (18–56 mm), based mainly on the occurrence of Coprosma, as opposed to a wetter month of 270–320 mm. Additionally, the same values are generated for MPwarm (145–175 mm) as in the Warkalli beds (Ramanujam, 1987). Furthermore, the paleoflora from two outcrops (Kalarakod and Nirkunnam) of the Warkalli beds in Alleppey district (central Kerala) was analyzed (Rao, 1995). Due to the presence of only 18 significant NLR taxa in this list, the reconstructed paleoclimate estimates show much larger intervals (Table 2). Nevertheless, trends are similar to the others from the Wakalli beds and the Quilon beds and point towards a moderate seasonality of temperature and precipitation. 4.2.3. Climate of South Indian east coast In addition to the rich paleofloras on the west coast of southern India, there are outcrops in the eastern parts (Tamil Nadu province) of similar age from the Neyveli Formation and the overlying Cuddalore Formation. Their dating is still problematic; however, they are rich in paleobotanical fossils (e.g., Ramanujam, 1987; Mandaokar and Mukherjee, 2012). For comparison, two studies (Singh et al., 1992; Mandaokar and Mukherjee, 2012) were selected and analyzed with the CA, which are dated to the early to middle Miocene. The CA resulted in climatic values, which resemble that from the west coast. More precise climatic data were generated for the presumable early Miocene locality of the Neyveli Formation (Singh et al., 1992), where the results of palynoflora and fossil wood were combined. Because of the presence of additional wood elements Mesua (Clusiaceae), Hopea (Dipterocarpaceae), Phyllanthus (Euphorbiaceae), and Diopyros (Ebenaceae), narrower climatic intervals could be estimated, along with the determination of pollen of Cupania (Rubiaceae). More significant is the presence of Pterocarya and Nyssa, in contrast with the clear tropical elements of the assemblage, which lead to MAT of 23.9–24.4 ◦ C, CMT of 19.4–23.5 ◦ C and WMT of 27.5–27.8 ◦ C. More important is the comparable seasonality in rainfall: the driest month is as well expressed as in the west (18–56 mm), as is the wettest month (270–320 mm). Furthermore, precipitation for the warmest month points to slightly wetter conditions (163–175 mm) due to the presence of Hopea. 127 Comparable to the climatic estimates of the west is the MAP interval of 1748–1864 mm. A paleoflora from the overlying Cuddalore Formation was described from the Panruti locality, which is dated as early Miocene by Mandaokar and Mukherjee (2012). Due to its lower diversity, the coexistence intervals are rather wide (Table 2) but the general pattern is in agreement with the other data. 5. Discussion 5.1. Depositional environment The deposits of the Varkala cliff section show different environments linked to minor sea-level changes (Reuter et al., 2013). Basically, the phases of interlamination of sand and organic-rich muds, lenticular, flaser and cross-bedding represent the deepest deposits, indicating shallow water deposits in the intertidal zone (Reineck, 1979; Reading and Collinson, 1996). There, the sanddominated facies point to sand flat environments near the low water line. Otherwise, the clay dominated facies is typical for higher intertidal mud flats that form near the high water line (Reineck, 1979). Here, the structureless sandy, silty clays possibly originated from the homogenization of sand, silt and clay layers through intense burrowing and/or plant rooting. In these horizons, coastal vegetation is documented by dense, deep-reaching rootlets highly reminiscent of their size and form of mangrove roots (Fig. 3F). Where sand is missing and mangrove roots are rare, sedimentation indicates a coastal backswamp environment similar to today. In beds 13 and 18, the coarse-grained siliciclastic deposits with plane, gently seaward dipping laminae of well sorted coarse sand and fine gravel, were formed in the swash zone (Reinson, 1984) extending from the limit of run-down to the limit of runup of the wave action. Accordingly, the steep break of slope at the upper surface of bed 18 is interpreted as the face of a beach berm (Reading and Collinson, 1996). Further, the red horizon in bed 13 represents a paleosol, likewise the phase of lowest sea level. In the whole Varkala cliff section, most prominent bioturbation was caused by crabs which produced the vertical burrows with chevron pattern (Melchor et al., 2010) and further indicates mangroves, tidal flats and beaches, where they typically exist (Eisma, 1998). Associated Diplocraterion ichnofossils are made by the amphipod Corophium, which lives in the intertidal zone (Yeo and Risk, 1981). Thus, the sedimentary record clearly suggests a coastal environment, representing various settings from the intertidal zone near the low water line to the vegetated coast and its backswamps. 5.2. Vegetation reconstruction at Varkala cliff section Because pollen determination was limited mainly to family level, the vegetation reconstruction can only outline the past plant community. However, high similarities to the vegetation of Kerala today are obvious. The presence of many ancient Author's personal copy 128 A.K. Kern et al. / Palaeoworld 22 (2013) 119–132 rainforest plants is possibly linked with the fact that the Western Ghats were a retreat area during the Pleistocene (Farooqui et al., 2010). Therefore, a comparison with modern taxa as the potential representatives of the recorded families was used to draw a generalized picture of the paleovegetation. A coastal setting is already evident from the sedimentology (discussed above in 5.1.). The coastline was fringed by mangroves, indicated by the imprint of their breathing roots (e.g., beds 11, 14; Fig. 2). Although no clearly identifiable macroscopic plant material is preserved in these horizons, the abundance, uniformity, and thinness are highly reminiscent of the roots of Avicenniaceae in particular (Fig. 3F). Mangroves are the most reported vegetation type indicated by the pollen analysis. Rhizophoraceae are the abundant plants in the Indian mangroves today (Thanikaimoni, 1987; Maria and Sridhar, 2002); they were, however, diverse, but less important in the past. Rhizophoraceae produce a peak in pollen grains (Thanikaimoni, 1987), which results in an account above 80% in paleo-records (e.g., Engelhart et al., 2007) but occurs only up to 13.5% in our samples. Along the shoreline, the palm genus Nypa is absent in Kerala today, and the mangrove apple Sonneratia is exclusively related with brackish and marine habitats along with mangrove habitats as is the Meliaceae genus Xylocarpus (Raju, 2003). Avicenniaceae, represented only by the genus Avicennia worldwide today, accounts for up to 7% in the samples. Despite this low count, Avicennia is one of the most abundant mangrove plants during the deposition of the Warkalli beds because Avicennia produces a very low number of pollen (Thanikaimoni, 1987). The documented Euphorbiaceae, Myrsiaceae, and Plumbaginaceae might also have been the components of these coastal mangroves (Thanikaimoni, 1987; Gopal and Krishnamurthy, 1993). Backswamps are formed adjacent to the main mangrove belt, where plants adapted to less salty soil lived, such as Acanthaceae, Compretaceae, Malvaceae, Lecythaceae, Brownlowia (Malvaceae) and Caesalpinioideae. Typical modern representatives in Kerala are Lumnitzera (Compretaceae), Hibiscus (Malvaceae) and Phoenix (Arecaceae) along with Suaeda (Chenopodiaceae), and Urochondra and Porteresia (Poaceae) (Gopal and Krishnamurthy, 1993; Kokkal et al., 2008). Additionally, plants restricted to freshwater habitats are present in our investigated samples. Plants like Typha (Typhaceae) or Myriophyllum (Haloragaceae) may have lived in ponds or slow flowing rivers behind the coastline and its backswamps. This may be similar to today, where the regional setting of Kerala is characterized by a long coastline, covered by far distributed mangroves in connection with wide backswamp vegetation due to the rivers flowing westwards to the sea from the Western Ghats (Farooqui et al., 2010). Further in the hinterland, including the low elevation area along the mountain chain, a mainly evergreen forest was growing, most likely comparable to the tropical evergreen to deciduous low elevation forest in this area today (Farooqui et al., 2010). Although rather rare in the pollen-spectra, the nowadays important tropical canopy tree Dipterocarpaceae (Barboni and Bonnefille, 2001) seems to have been a typical element of these forests based on the very abundant amber found partly sill inside the fossil woods. Resin is ubiquitous in all sections of the Warkalli Group and was formed by members of the Dipterocarpaceae family according to Dutta et al. (2010) and our studies (N. Vávra, pers. comm., 2010). Additional assumptions of the paleovegetation are based only on the occurrence of family-level taxa, which comprise genera typical for tropical rainforests in the Asian paleotropics today such as Fabaceae (Kingiodendron), Clusiaceae (Calophyllum, Garcinia, Mesua), Sapotaceae (Palaquium), Euphorbiaceae (Drypetes, Fahrenheitia, Macaranga, Mallotus), Myristicaceae (Myristica, Knema), Myrtaceae (Syzygium), Anacardiaceae (Mangifera, Holigarna), and Meliaceae (Barboni and Bonnefille, 2001). Undergrowing plants could be considered according to recent distribution of Rubiaceae (Ixora, Psychotria), Fabaceae (Humboldia) or Meliaceae (Agaia) (Barboni and Bonnefille, 2001). Some taxa, such as Sapotaceae (Palaquium), Clusiaceae (Mesua, Colphyllum), Meliaceae (Aglaia), Euphorbiaceae (Drypetes), along with Dipterocarpaceae (Hopea, Vateria) and several species of Anacardiaceae, Myristicaceae and Myrtaceae, also occur higher up the mountains as part of the mid elevation vegetation (650–1400 masl) next to an increasing number of deciduous taxa. Pollen of Calamus, Bombacaceae (Bombax), Anacardiaceae (Buchanania), Lecythidaceae (Careya), Combretaceae (Terminalia), Fabaceae (Dalbergia, Cassia), and Mimosaceae (Acacia, Albizia) in the fossil samples might represent this mid elevation vegetation. A significant difference between the early–middle Miocene and today is the missing of gymnosperms, which did not become the common constituents of the flora in Kerala before the Pleistocene (Bonnefille et al., 1999; Barboni and Bonnefille, 2001; Farooqui et al., 2010; Paul and George, 2010). Nevertheless, Araucariaceae fossils were already rarely reported as the only gymnosperm in the early Miocene (Bande and Prakash, 1986) and Pinaceae and Podocarpaceae were already present in northeastern India (Mandaokar, 2002). As illustrated in Fig. 4, the vegetation within the Varkala cliff section is not significantly changing. Although mangroves are dominating in all the samples, the true mangrove and the mangrove associated plants are less abundant in samples 1, 3, and 12. However, this may be due more likely to taphonomy and the shifts in the position of the shoreline as to an essential change in the paleoflora. 5.3. Comparison of the different climatic data Although different localities in the western and the eastern parts of South India were analyzed herein, these similarities of the vegetation were already discussed in many of the previous papers (Rao and Ramanujam, 1975; Ramanujam, 1987; Singh et al., 1992). Variations were interpreted to be caused by sample position in relation to their distance to sea as these authors distinguished more marine influenced mangrove vegetation from more freshwater dominated swamp vegetation. However, no variation in the tropical vegetation over time between the early Miocene and the middle Miocene could be found. Our climatic estimate by the plant-based CA supports this assumption with almost identical climatic intervals. Author's personal copy A.K. Kern et al. / Palaeoworld 22 (2013) 119–132 All MAT intervals (except for the Warkalli beds) are relatively wide and would overlap, as do the CMT and WMT data (Table 2). To sum up, the temperature varied approximately between 19.4 ◦ C and 23.8 ◦ C in the coldest month and rose up to approximately 27.5–28.1 ◦ C in the warm month. Given all MAT data, a wide interval between 22.2 ◦ C and 27.5 ◦ C may be considered as the most reliable estimate. Further, the similarity in the rainfall values is of high relevance; the reconstruction for the mean annual precipitation is between 1747 mm and 2099 mm. A wetter and drier season was established with a difference of MPdry 18–83 mm and MPwet 270–358 mm for all samples, respectively. The higher amount of rainfall may have been closer to the warmer season as the mean precipitation of the warmest month reaches up to 175 mm, although a clear connection of temperature and rainfall is beyond these data. In comparison with the recent climate in southern India, significant differences emerge. Although temperature appears to be similar, there is a significant difference within the distribution of rainfall. Southern India’s precipitation pattern is linked to the Indian Monsoon (South Asian Monsoon), which originates in the equatorial Indian Ocean and enters India from the Arabian Sea (Wang et al., 2003). The Indian Monsoon causes the highest rainfall rates in Southern Asia (Gopal and Krishnamurthy, 1993) with more than 5000 mm per year and 800 mm per month recorded in Kerala (Lieth et al., 1999; Barboni and Bonnefille, 2001; Pal and Al-Tabbaa, 2009). There, the mountain chain of the Western Ghats is located less than 200 km away from the sea and consequently acts as a rain barrier due to its peaks up to more than 2500 m high (Barboni and Bonnefille, 2001). Along the west coast, the rainfall season ranges from June to September (Lieth et al., 1999), which represents 67.9% of the annual precipitation (Krishnakumar et al., 2009). Although average levels of rainfall along the west coast are reported as high, less rainfall occurs around the southern tip of the Indian Peninsula, which does not receive precipitation during this first phase of the Indian Monsoon (Reuter et al., 2013; e.g., Trivandrum; MAP = 1736 mm; www.worldclimate.com). The climate in the southern part and the south-eastern part of India is also dominated by the Indian Monsoon, but the rainy season is during October to December, when the center of main precipitation is retreating to the southwest (post-monsoon-season) triggered by the formation of a low pressure zone above the Bay of Bengal during late autumn (Kripalani and Kumar, 2004). As a consequence, strikingly lower mean annual precipitation occurs in the eastern part with values between approximately 800 mm and 1500 mm (Marlange and Meher-Himji, 1965; Venkateswaran and Parthasarathy, 2003), often with irregularities in the area of main rainfall (Blasco and Legris, 1972). This further causes differences in the vegetation in the east. Although it is still tropical with mangroves along the coast, the main vegetation is labeled as tropical dry evergreen forest (Champion and Seth, 1968) or seasonal dry evergreen forest (Venkateswaran and Parthasarathy, 2003). This is linked to a lower plant diversity than in the west (approximately 30 families) and a higher amount of deciduous trees (Venkateswaran and Parthasarathy, 2003 and references therein). 129 The contrast, however, is not present in the paleovegetation (Rao and Ramanujam, 1975; Ramanujam, 1987; Singh et al., 1992; Mandaokar and Mukherjee, 2012) as well as in our results of the CA. Few additional temperate to subtropical elements occurred in the early Miocene eastern Neyveli sample (Ulmaceae, Pterocarya, Engelhardia), but they do not cause a shift in the climatic intervals as they are accompanied by a rich tropical flora (also excluding the fossil wood taxa). The paleovegetation of the Miocene East India includes several elements, which are not distributed in this area today, such as highly diverse Dipterocarpaceae, Arecaceae, Symplocaceae or Sapotaceae (Venkateswaran and Parthasarathy, 2003; Sukumaran et al., 2008). This may support the idea of a greater difference in the climate of south-eastern India than in the west compared to today and a more uniform vegetation and climate in the early and middle Miocene of South India. To summarize these data focusing solely on CA precipitation values (lower MAT and MPwet along the west coast), it could be assumed the Indian Monsoon was less prominent during this period. Today’s rainfall barrier of the Western Ghats did already exist (e.g., Beane et al., 1986; Pande, 2002), which therefore cannot explain the similarity of rainfall patterns. Clift et al. (2008) described the middle Miocene Climatic Optimum as a weak phase of the Indian Monsoon, which would fit the interpretation of our climatic data for the western localities. However, this would possibly not lead to these wetter conditions in the east. Furthermore, we cannot observe variations from the early Miocene localities and the one from middle Miocene. Next to the missing precipitation gradient from east to west, the CA data also do not show any change from the south to the north. The more southern position of India at this time may result in the missing of monsoonal rainfall at the Varkala cliff section (Reuter et al., 2013), and cannot explain the similarity of all the sites along the Kerala shoreline. One problem in comparing all these localities may be the low taxonomic resolution of plant determination at family level, which makes it impossible to detect genera/species adapted to a climate with a dry season. Furthermore, mapping of tropical taxa and the lower coverage of meteorological stations in these areas could cause less accurate climatic reconstructions with the CA. Additionally, uncertainties in the dating of these paleofloras could obscure the variations of the Miocene Indian Monsoon intensities. Thus, these contradictions prevent us from discussing the evolution of the early Indian Monsoon in greater detail. Further studies in whole India and the surrounding countries as well as climatic reconstructions based on other paleoclimate proxies need to be done to draw distinct conclusions about its behavior and intensity in the early and middle Miocene. 6. Conclusion The coastal deposits of the Warkalli Group are the only onshore opportunity to gather data on paleoenvironment and paleoclimate in southwestern India during the late early Miocene to middle Miocene. The rich palynoflora at the Varkala cliff section consists of the plants that are still present in the recent Author's personal copy 130 A.K. Kern et al. / Palaeoworld 22 (2013) 119–132 flora in the tropical south and southeastern Asian region. Sedimentological data show a coastal marine setting with intensely bioturbated shore sands, which shift to intertidal mud flats and backswamp deposits. This accounts for the rapid successions of small scale relative sea level fluctuations. Mangrove-fringes are reconstructed by pollen and are indicated by certain taxa of mainly Avicennia, Nypa, Sonneratia, Xylocarpus and Rhizophoraceae along with elements of the families Myrsinaceae, Combretaceae, Acanthaceae, and Malvaceae. Adjacent to these were backswamps, followed by a growing evergreen forest, which were merged into an increasing number of deciduous plants at higher elevations of the already existing mountains in the Western Ghats. Despite many comparable elements in the fossil and extant plant assemblages, some critical differences exist by reconstructing the paleoclimatological conditions for the Varkala cliff section. Based on the CA data of the samples the mean annual precipitation did not exceed 1958 mm, which is distinctly lower than the average of ∼3000 mm in Kerala Province today. Seasonality in rainfall was contrasted by the wettest month of 225–358 mm and a dry month with 18–83 mm precipitation. Values for the rainfall during the warmest month (114–175 mm) suggest slightly more humid conditions during the warmer month, which is quite different from the climate of this area today. However, estimates for temperatures are similar (mean annual temperature 22.2–27.5 ◦ C; coldest month temperature 20.6–23.6 ◦ C; warmest month temperature 27.5–28.1 ◦ C). These data slightly vary from previously published data by Reuter et al. (2013) due to updates in the climatic data base used (www.palaeoflora.de). Comparison between different localities along the coastline of Kerala of early to middle Miocene age showed a high similarity in the regional paleoflora (Rao and Ramanujam, 1975; Ramanujam, 1987; Singh et al., 1992; Rao, 1995; Mukherjee, 2012) and likewise the results of the plant-based Coexistence Approach. Furthermore, the vegetation of SW India is in good accordance with those in SE India of similar age. The botanical and climatic similarity between these areas suggests a more uniform climate than today, but also mirrors the complexity of the evolution of tropical areas towards the botanical and climatic modern situation. However, the stratigraphic uncertainties between the outcrops may obscure the climatic reconstructions. Besides, the determination of the paleoflora was limited mainly to the family level, which does not allow deciphering variations between the SE and SW Indian vegetation in greater detail. Acknowledgments The study was supported by the FWF-grants P-18189-N10 and P-21414-B16. We acknowledge the Geological Survey of Vienna for providing the lab for sample preparation and thank Prof. Dr. Norbert Vávra for chemical amber analyses. Gonzalo Jimenez-Moreno helped with pollen identification and provided a copy of Thanikimaimoni “Mangrove Palynology”, and Torsten Utescher and Manju Banerjee for their critical and constructive comments in their reviews. This publication is contributing to the NECLIME network. Appendices A and B. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.palwor.2013.10.001. References Badve, R.M., Sakurkar, C.V., 2003. On the disappearance of palm genus Nypa from the west coast with its present status in the Indian subcontinent. Current Science 85, 1407–1409. Baker, W.J., Couvreur, L.P.J., 2012. Biogeography and distribution patterns of Southeast Asian palms. In: Gower, D.J., et al. (Eds.), Biotic Evolution and Environmental Change in Southeast Asia. Cambridge University Press, pp. 164–190. Bande, M.B., Prakash, U., 1986. The Tertiary flora of Southeast Asia with remarks in its palaeoenvironment and phytogeography of the Indo-Malayan region. Review of Palaeobotany and Palynology 49, 203–233. Barboni, D., Bonnefille, R., 2001. Precipitation signal in pollen rain from tropical forests, South India. Review of Palaeobotany and Palynology 114, 239–258. Beane, J.E., Turner, C.A., Hooper, P.R., Subbarao, K.V., Walsh, J.N., 1986. Stratigraphy, composition and form of the Deccan Basalts, Western Ghats, India. Bulletin of Volcanology 48, 61–83. Blasco, F., Legris, P., 1972. Dry evergreen forest of Point Calimere and Marakanum. Journal of the Bombay Natural History Society 70, 279–294. Bonnefille, R., Anupama, K., Barboni, D., Pascal, J., Prasad, S., Sutra, J.P., 1999. Modern pollen spectra from tropical South India and Sri Lanka: altitudinal distribution. Journal of Biogeography 26, 1255–1280. Campanile, D., Nambiar, C.G., Bishop, P., Widdowson, M., Brown, R., 2008. Sedimentation record in the Kankan-Kerala Basin: implications for the evolution of the Western Ghats and the Western Indian passive margin. Basin Research 20, 3–22. Champion, H.G., Seth, V.K., 1968. A Revised Survey of the Forest Types of India. Government of India, Delhi, 404 pp. Clift, P.D., Hodges, K.V., Heslop, D., Hanningan, R., Van Long, H., Calves, G., 2008. Correlation of Himalayan exhumation rates and Asian monsoon intensity. Nature Geoscience 1, 875–880. Dove, J.L., 2010. Australian Palms: Biogeography, Ecology and Systematics. Csiro Publishing, Collingwood, 292 pp. Dutta, S., Mallick, M., Mathews, R.P., Mann, U., Greenwood, P.F., Saxena, R., 2010. Chemical composition and palaeobotanical origin of Miocene resins from Kerala-Konkan coast, western India. Journal of Earth System Science 199, 711–716. Eisma, D., 1998. Intertidal Deposits: River Mouths, Tidal Flats, and Coastal Lagoons. CRC Press, Boca Raton, 525 pp. Ellison, J., Koedam, N.E., Wang, Y., Primavera, J., Jin Eong, O., Wan-Hong Yong, J., Ngoc Nam, V., 2010. Nypa fruticans. In: IUCN 2012. IUCN Red List of Threatened Species. Version 2012.2, <www.iucnredlist.org>. Downloaded on 17 April 2013. Engelhart, S.E., Horton, B.P., Roberts, D.H., Bryant, C.L., Corbett, D.R., 2007. Mangrove pollen of Indonesia and its suitability as sea-level indicator. Marine Geology 242, 65–81. Erdtman, G., 1954. An Introduction to Pollen Analysis. Chronica Botanica Company, Waltham, Massachusetts, 239 pp. Farooqui, A., Ray, J.G., Farooqui, S.A., Tiwari, R.K., Khan, Z.A., 2010. Tropical rainforest vegetation, climate and sea level during the Pleistocene in Kerala, India. Quaternary International 213, 2–11. Gopal, B., Krishnamurthy, K., 1993. Wetlands of South Asia. In: Whigham, D.F., Dykyjová, D., Hejný, S. (Eds.), Wetlands of the World I: Inventory, Ecology and Management. Kluwer Academic Publishers, Dordrecht, pp. 345–414. Grimm, E.C., 2004. Tilia and TG View Version 2.0.2. Illinois State Museum, Research and Collector Center. Guo, Z., Ruddiman, W.F., Hao, Q.Z., Wu, H.B., Qian, Y.S., Zhu, R.X., Peng, S.Z., Wei, J.J., Yuan, B.Y., Liu, T.S., 2002. Onset of Asian desertification by 22 Ma ago inferred from loess deposits in China. Nature 416, 159–163. Hemp, A., 2011. Altitudinal zonation and diversity patterns in the forests of Mount Kilimanjaro, Tanzania. In: Bruijnzeel, L.A., Scatena, F.N., Hamilton, Author's personal copy A.K. Kern et al. / Palaeoworld 22 (2013) 119–132 L.S. (Eds.), Tropical Montane Cloud Forests: Science for Conservation and Management. Cambridge University Press, Cambridge, pp. 134–141. Hitimana, J., Legilisho Kiyiapi, J., Thairu Njunge, J., 2004. Forest structure characteristics in disturbed and undisturbed sites of Mt. Elgon Moist Lower Montane Forest, western Kenya. Forest Ecology and Management 194, 269–291. Johnson, D., 1998. Oncosperma fasciculatum. In: IUCN 2012. IUCN Red List of Threatened Species. Version 2012.2, <www.iucnredlist.org>. Downloaded on 30 April 2013. Klaus, W., 1987. Einführung in die Paläobotanik, Band 1. Franz Deuticke, Wien, 295 pp. Kokkal, K., Harinarayanan, P., Sabu, K.K., 2008. Wetlands of Kerala. Proceedings of Taal 2007: The 12th World Lake Conference, pp. 1889–1893. Kripalani, R.H., Kumar, P., 2004. Northeast monsoon rainfall variability over South Peninsular India vis-à-vis the Indian Ocean dipole mode. International Journal of Climatology 24, 1267–1282. Krishnakumar, K.N., Prasada Rao, G.S.L.H.V., Gopakumar, C.S., 2009. Rainfall trends in twentieth century over Kerala, India. Atmospheric Environment 43, 1940–1944. Lieth, H., Berlekamp, J., Fuest, S., Riediger, S., 1999. Climate Diagram World Atlas, 1st edition. Backhuys Publishers, Leiden, ISBN 90-5782-031-5 [CDROM]. Ling, K.H., Wang, Y., Poon, W.S., Shaw, P.C., But, P.P.H., 2009. The relationship of Fagaropsis and Luvunga in Rutaceae. Taiwania 54, 338–342. Madulid, D., 1998. Oncosperma platyphyllum. In: IUCN 2012. IUCN Red List of Threatened Species. Version 2012.2, <www.iucnredlist.org>. Downloaded on 30 April 2013. Mandaokar, B.D., 2002. Palynoflora from the Keifang Formation (Early Miocene) Mizoram, India and its environmental significance. Journal of the Palaeontological Society of India 47, 77–83. Mandaokar, B.D., Mukherjee, D., 2012. Palynological investigation of Early Miocene sediments exposed at Panruti, Cuddalore district, Tamil Nadu, India. International Journal of Geology, Earth and Environmental Sciences 2, 157–175. Maria, G.L., Sridhar, K.R., 2002. Richness and diversity of filamentous fungi on woody litter of mangroves along the west coast of India. Current Science 83, 1573–1580. Marlange, M., Meher-Himji, V.M., 1965. Phytosociological studies in the Pondicherry region. Journal of the Indian Botanical Society 44, 167–182. Martinetto, E., 2001. The role of central Italy as a centre of refuge for thermophilous plants in the late Cenozoic. Acta Palaeobotanica 41, 299–319. Mehrotra, R.C., Tiwari, R.P., Mazumder, B.I., 2003. Nypa megafossils from the Tertiary sediments of Northeast India. Geobios 36, 83–92. Melchor, R.N., Genise, J.F., Farina, J.L., Sánchez, M.V., Sarzetti, L., Visconti, G., 2010. Large striated burrows from fluvial deposits of the Neogene Vinchina Formation, La Rioja, Argentinia: a crab origin suggested by neoichnology and sedimentology. Palaeogeography, Palaeoclimatology, Palaeoecology 291, 400–418. Misra, B.K., Singh, A., Ramanujam, C.G.K., 1996. Trilatiporate pollen from Indian Palaeogene and Neogene sequences: evolution, migration and continental drift. Review of Palaeobotany and Palynology 91, 331–352. Mosburgger, V., Utescher, T., 1997. The coexistence approach — a method for quantitative reconstructions of Tertiary terrestrial palaeoclimate data using plant fossils. Palaeogeography, Palaeoclimatology, Palaeoecology 134, 61–86. Mukherjee, D., 2012. Facultative fungal remains from Miocene lignite coal of Neyveli Tamil Nadu, India. International Journal of Geology, Earth and Environmental Sciences 2, 1–15. Munishi, L.K., 2007. The distribution and diversity of tree resources outside forest in southern side of Mount Kilimanjaro. Discovery and Innovation 19, 36–44. Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B., Kent, J., 2000. Biodiversity hotspots for conservation priorities. Nature 403, 853–858. Pal, I., Al-Tabbaa, A., 2009. Trends in seasonal precipitation extremis — an indicator of ‘climate change’ in Kerala, India. Journal of Hydrology 367, 62–69. 131 Pande, K., 2002. Age and duration of the Deccan Traps, India: a review of radiometric and palaeomagnetic constraints. Proceedings of the Indian Academy of Sciences 111, 115–123. Paul, J., George, K.V., 2010. Studies on riverine flora of Pamba river basin, Kerala. Nature Proceedings, http://dx.doi.org/10.1038/npre.2010.5135.1 Partomihardjo, T., Mirmanto, E., Riswan, S., Whittaker, R.J., 1992. Ecology and distribution of Nibung (Oncosperma tigillarium) within the Krakatau Islands, Indonesia. Principes 36, 7–17. Quan, C., Liu, Y.S., Utescher, T., 2012. Paleogene temperature gradient, seasonal variation and climate evolution of northeast China. Palaeogeography, Palaeoclimatology, Palaeoecology 313–314, 150–161. Raha, P.K., 1996. A revision of the stratigraphic sequence of coastal sedimentary basin of Kerala. 15th Indian Colloquium of Micropalaeontology and Stratigraphy, Dehra Dun, pp. 805–810. Raju, J.S.S.N., 2003. Xylocarpus (Meliaceae): a less-known mangrove taxon of the Godavari estuary, India. Current Science 84, 879–881. Ramanujam, C.G.K., 1982. Tertiary palynology and palynostratigraphy of Southern India. The Palaeontological Society of India, Special Publication 1, 57–64. Ramanujam, C.G.K., 1987. Palynology of the Neogene Warkalli Beds of Kerala State in South India. Journal of the Palaeontological Society of India 32, 26–46. Rao, A.N., Lee, Y.K., 1970. Studies on Singapore pollen. Pacific Science 24, 255–268. Rao, K.P., Ramanujam, C.G.K., 1975. A palynological approach to the study of Quilon Beds of Kerala State in South India. Current Science 44, 730–732. Rao, M.R., 1995. Palynostratigraphic zonation and correlation of the Eocene–Early Miocene sequence in Alleppy district, Kerala, India. Review of Palaeobotany and Palynology 86, 325–348. Reading, H.G., Collinson, J.D., 1996. Clastic coasts. In: Reading, H.G. (Ed.), Sedimentary Environments: Processes, Facies and Stratigraphy. Blackwell Science, Oxford, pp. 154–231. Reineck, H.E., 1979. German North Sea Tidal Flats. In: Ginsburg, R.N. (Ed.), Tidal Deposits, A Casebook of Recent Examples and Fossil Counterparts. Springer-Verlag, Berlin, Heidelberg, New York, pp. 5–12. Reinson, G.E., 1984. Barrier island and associated strand-plain systems. In: Walker, R.G. (Ed.), Facies Models. Geoscience Canada Reprint Series 1, 119–140. Reuter, M., Piller, W.E., Harzhauser, M., Kroh, A., Rögl, F., Coric, S., 2011. The Quilon Limestone, Kerala Basin, India: an archive for Miocene Indo-Pacific seagrass beds. Lethaia 44, 76–86. Reuter, M., Kern, A.K., Harzhauser, M., Kroh, A., Piller, W.E., 2013. Global warming and South Indian monsoon rainfall — lessons from the MidMiocene. Gondwana Research 23, 1172–1177. Saxena, R.K., 1982. Taxonomic study of the polycolpate pollen grains from the Indian Tertiary sediments with special reference to nomenclature. Review of Palaeobotany and Palynology 37, 283–315. Singh, A., Misraa, B.K., Singh, B.D., Navalea, G.K.B., 1992. The Neyveli lignite deposits (Cauvery basin), India: organic composition, age and depositional pattern. International Journal of Coal Geology 1–2, 45–97. Srivastava, G., Spicer, R.A., Spicer, T.E.V., Yang, J., Kumar, M., Mehrotra, R., Mehrotra, N., 2012. Megaflora and palaeoclimate of a Late Oligocene tropical delta, Makum Coalfield, Assam: evidence for the early development of the South Asia Monsoon. Palaeogeography, Palaeoclimatology, Palaeoecology 342–343, 130–142. Sukumaran, S., Jeeva, S., Raj, A.D.S., Kannan, D., 2008. Floristic diversity conservation status and economic value of miniature sacred groves in Kanyakumari district, Tamil Nadu, Southern Peninsular India. Turkish Journal of Botany 32, 185–198. Sun, X., Wang, P., 2005. How old is the Asian monsoon system? — Palaeobotanical records from China. Palaeogeography, Palaeoclimatology, Palaeoecology 222, 181–222. Thanikaimoni, G., 1987. Mangrove Palynology. UNDP/UNESCO Regional Project on Training and Research on Mangrove Ecosystems, RAS/79/002 and the French Institute, Pondicherry, 100 pp. Vaidyanadhan, R., Ramakrishnan, M., 2008. Geology of India. Geological Society of India, Bangalore, 994 pp. Author's personal copy 132 A.K. Kern et al. / Palaeoworld 22 (2013) 119–132 Venkateswaran, R., Parthasarathy, N., 2003. Tropical dry evergreen forests in the Coromandel coast of India: structure, composition and human disturbance. Ecotropica 9, 45–58. Venkatachala, B.S., Rawat, M.S., 1972. Palynology of Tertiary sediments in the Cauvery Bas: I. Palaeocae–Eocene palynoflora from subsurface. In: Ghosh, A.K. (Ed.), Proceedings, Seminar on Paleopalynology and Indian Stratigraphy. University of Calcutta, Calcutta, pp. 292–334. Venkatachala, B.S., Rawat, M.S., 1973. Palynology of Tertiary sediments in the Cauvery Basin. 2. Oligocene–Miocene palynoflora from the subsurface. Palaeobotanist 20, 238–263. Wan, S., Li, A., Clift, P.D., Stuut, J.B.W., 2007. Development of the East Asian monsoon: mineralogical and sedimentological records in the northern South China Sea since 20 Ma. Palaeogeography, Palaeoclimatology, Palaeoecology 254, 561–582. Wang, B., Clemens, S.C., Lui, P., 2003. Contrasting the Indian and East Asian monsoons: implications on geologic timescales. Marine Geology 201, 5–21. Yeo, R.K., Risk, M.J., 1981. The sedimentology, stratigraphy, and preservation of intertidal deposits in the Minas Basin System, Bay of Fundy. Journal of Sedimentary Petrology 51, 245–260.