peatland system

Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

Contents lists available at ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology journal homepage: www.elsevier.com/locate/palaeo

Cyclic floral succession and fire in a Cenozoic wetland/peatland system Vera A. Korasidis a,⁎, Malcolm W. Wallace a, Barbara E. Wagstaff a, Guy R. Holdgate a, Anne-Marie P. Tosolini a, Ben Jansen b a b

School of Earth Sciences, The University of Melbourne, Victoria 3010, Australia GHD, 5 Church St, Traralgon, Victoria 3844, Australia

a r t i c l e

i n f o

Article history: Received 22 January 2016 Received in revised form 21 August 2016 Accepted 25 August 2016 Available online 30 August 2016 Keywords: Lithotype Palynology Vegetation succession Latrobe Valley Brown coal Oligocene-Miocene

a b s t r a c t The cyclic succession of brown coals in the Latrobe Valley, Gippsland Basin, Australia, records an exceptional floral and charcoal record from the Late Oligocene to Middle Miocene. New palynological, geological and charcoal data are consistent with existing colourimetry, carbon isotope, and organic geochemical and palaeobotanical data, indicating that the repeated lithotype cycles represent relative drying (terrestrialization). Based on this detailed palynological study, the vegetation succession within the Latrobe Valley peatlands is interpreted to have begun with a fire-prone emergent marsh of bulrushes (Typhaceae), which grades landward into a fire-prone meadow marsh of rushes (Restionaceae), heaths (Ericaceae) and coral-ferns (Gleicheniaceae). This marsh environment then developed into a forested bog, with gymnosperms (e.g. the Podocarpaceae Dacrycarpus and Dacrydium) as the dominant trees, until an ombrogenous forest bog developed, predominantly consisting of angiosperms (e.g. Nothofagus, Quintinia). The similarity between vegetation successions in New Zealand and the lightening-upwards cycles from the Latrobe Valley coals suggests that New Zealand's modern vegetation communities represent a floral analogue for the successions preserved in the Latrobe Valley coals. High abundances of micro and macro charcoal recorded in the darker lithotypes, within the lithotype cycles of the M1B and M2A seams, suggest that the Latrobe Valley peatlands were subject to repeated fires during the Late Oligocene to Early Miocene. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Repeated floral successions with abundant charcoal are preserved in the lithotype cycles of the Latrobe Valley brown coals. However, the origin of these lithotype cycles remains controversial with studies from different disciplines producing contradictory models termed the drydark and dry-light models (Holdgate et al., 2014). The same two contradictory models have also been proposed for the German brown coals. The dry-dark model proposed that dark lithotypes were produced in the driest facies, while light lithotypes were deposited in either very wet or open water environments (Teichmüller, 1958; Teichmüller, 1989; Luly et al., 1980; Kershaw and Sluiter, 1982; Sluiter and Kershaw, 1982; Finotello and Johns, 1986; Blackburn and Sluiter, 1994). In contrast, the dry-light model proposed that light lithotypes were deposited in the driest facies, while dark lithotypes were deposited in the wettest facies (Hagemann and Hollerbach, 1980; Hagemann and Wolf, 1987; Anderson and Mackay, 1990; Holdgate et al., 1995; Holdgate et al., 2014). These lithotypes are arranged in cycles, which

⁎ Corresponding author at: School of Earth Sciences, The University of Melbourne, Victoria 3010, Australia. E-mail address: [email protected] (V.A. Korasidis).

http://dx.doi.org/10.1016/j.palaeo.2016.08.030 0031-0182/© 2016 Elsevier B.V. All rights reserved.

are generally attributed to being caused by terrestrialization (dryingupwards cycles) (e.g. Muller et al., 2003; Frank and Bend, 2004). This paper presents a more detailed depositional model for the Morwell 1B seam of the Latrobe Valley based on new palynological, geological and charcoal data, in conjunction with a re-analysis of preexisting colourimetry, carbon isotope, geochemical and palaeobotanical (macrofossils and palynology) data. Comparisons are also drawn from studies undertaken on modern vegetation successions in New Zealand, which appear to be similar, in terms of floral composition (at genera level), to the Latrobe Valley floras and successions. The flora and charcoal records from the Late Oligocene to Early Miocene also provide insights into the fire history of Australia's vegetation (Scott et al., 2000). 2. Geological setting The Gippsland Basin, in south-eastern Australia, records a sedimentary record from the Cretaceous to Recent and is unique with regard to the scale and size of its contained brown coal and petroleum energy resources (Smith, 1982). The Cenozoic onshore segment of the basin is dominated by terrestrial sediments of the Latrobe Valley Group (Figs. 1,2). The main coal-bearing sequences in the Latrobe Valley Group, of the onshore Gippsland Basin are, in stratigraphic order, the Traralgon

238

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

Fig. 1. Location of the Gippsland Basin and Latrobe Valley. Modified from Holdgate et al. (2014).

Formation, the Morwell Formation and the Yallourn Formation. The Morwell Formation, the focus of this paper, is divided into two major coal seams – the Morwell 1 (M1) and Morwell 2 (M2) seams, which in the eastern half of the Latrobe Valley Depression are further subdivided into the M1A, M1B, M2A, M2B and M2C seams. The coalbearing Latrobe Valley Depression hosts several coal seams N 100 m thick, constituting in total, the thickest brown coal succession in the world (Holdgate, 2003).

3. Lithotypes in the Latrobe Valley and lithotype models Lithotypes in the Latrobe Valley brown coals are defined on dry coal surfaces by colour, texture, gelification and weathering (George and

Mackay, 1991). Lithotypes become more defined on older and more weathered surfaces. Darker lithotypes in outcrop are characterized by high degrees of gelification, intense cracking and recessive weathering, relative to lighter lithotypes. Lighter lithotypes are characterized by resistant weathering, producing prominent beds with little or no cracking. George and Mackay (1991) defined 5 lithotypes: dark, medium dark, medium light, light and pale. Holdgate et al. (1995) defined a sixth lithotype, the laminated dark, characterized by the darkest colour and prominent lamination. Lithotypes occur as a series of cycles, the cycles ranging from 10 to 30 m in thickness and characterized by lightening-upward trends with prominent banding at a 1–3 m scale (Mackay et al., 1985; Holdgate et al., 2014). Cycle tops are commonly characterised by abrupt and unconformable boundaries with the overlying cycle (Mackay et al., 1985; Holdgate et al., 2014). Various models have been proposed to explain the formation of the lithotype cycles. Based on the soft (brown) coals of West Germany, Teichmüller (1958) first suggested a series of palaeoenvironments ranging from open water to dry forest, in a regular hydroseral series. Based on Teichmüller's (1958) model, Luly et al. (1980) proposed that the lithotypes in the brown coals of the Latrobe Valley ranged from open water (represented by the light lithotype), to increasingly drier swamp substrates (the darker lithotypes). Likewise, Kershaw et al. (1991) proposed that the darker lithotypes were deposited under the driest, most terrestrial conditions, and that the lighter lithotypes were formed in open water conditions. This model of darkening-up cycles is at odds with the observed lightening-upwards cycles first statistically documented by Mackay et al. (1985). Anderson and Mackay (1990) later contradicted these models and instead proposed that the dark lithotypes represented the wettest facies while the light lithotypes represented the driest. Anderson and Mackay (1990) suggested that the formation of the lithotypes took place in a series of ombrogenous peat bogs. These authors further suggested that peat dome development controlled the nature and extent of degradation and hence the nature of the lithotypes within the dome. Lighter lithotypes are suggested to form in the more oxic and drier (dry-light model), upper portions of the peat dome, while darker lithotypes form in the lower more anoxic, wetter facies (Anderson and Mackay, 1990). Holdgate et al. (2014) proposed a dry-light model with the lighteningupwards cycles representing relative drying (terrestrialization) upward for the Oligocene-Miocene brown coals in the Latrobe Valley. Holdgate et al. (2014) also integrated geological, geochemical and palynological data in order to interpret the major floral/ecological characteristics of the coal facies. More recently, Holdgate et al. (in press) have also applied this model to the German brown coals. In the Latrobe Valley, the cycles begin with a laminated dark or dark coal, the base of which can have high organic sulphur contents suggesting a marine-influence (Holdgate et al., 1995; Holdgate et al., 2014). 4. Methods Stratigraphic sections from the M1B and the M2A seams were measured on the southeastern face of Loy Yang Open Cut (Fig. 2). Coal samples were collected at 25 cm intervals based on the stratigraphic heights measured using a Jacob's staff. These samples were dried at 40 °C for 3 days and crushed to a grain size of 0.5–1 mm before quantitative colourimetry was performed using a Konica Minolta Chromameter CR-410 chromameter with the Hunter L, a, b colour scale. The three colour outputs – L (lightness), a (red) and b (yellow) were combined to produce a colour index using the formula as stated in Holdgate et al. (2014):

Fig. 2. Map of Loy Yang Open Cut Mine and the location of the M1B and M2A stratigraphic sections and bores (prefix LY). Modified from Holdgate et al. (2014)

Colour index ¼ 10ðL0 þ a0 þ b0Þ þ 100; where L ¼ ðL−16:966Þ=2:050; a0 ¼ ða−2:534Þ=0:604 and b0 ¼ ðb−4:421Þ=1:304:

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

This is the standard colour index used for colourimetry in the Latrobe Valley (e.g. Luly et al., 1980; HunterLab, 2007). At the beginning of each session the colourimeter was calibrated using a white calibration standard and an internal standard coal sample was repeatedly measured every 24 samples to determine precision. Lithotype colour range has been studied in detail by Attwood et al. (1984) and was found to be dependent on the grainsize of the crushed coal. For our procedures, we have assigned the following colour index values: laminated dark lithotype b25, dark − 25–49, medium dark − 50–84, medium light − 85–99, light − 100–149 and pale N 150. A detailed palynology study of the #64 m cycle (Fig. 3, stratigraphic numbering scheme from Holdgate et al., 1995) has been undertaken because this cycle has the most complete set of lithotypes, beginning with a laminated dark lithotype and gradually lightening upwards. The top of the cycle is truncated by an erosion surface at the #78 m cycle base and so the uppermost coals only reach medium dark colours. We have therefore included the top of the previous cycle (below #64 m) in order to analyse samples of lighter colour. 24 samples collected from the M1B seam at Loy Yang Open Cut were processed using palynological techniques including a 10% KOH solution, a weak Schulze solution (constructed by combining 320 ml of 35% HNO3 with 32 g of KCl), ammonia wash (5%), acetolysis, HCL (10%) and HF (48%). The process included the addition of exotic Lycopodium tablets to permit the calculation of pollen concentration. There was no fine sieving step included in the processing. Two hundred grains were counted per slide as this was determined to be statistically viable for the present study and for previous studies on the Latrobe Valley brown coals (Luly et al., 1980; Sluiter, 1984). In samples with limited spore-pollen, slides were counted to 100 grains. The slides were counted using a Zeiss AxioScope A1 microscope at × 400 magnification and the ×1000 magnification was used when necessary to identify distinguishing features of the spore-pollen. The modern affinities assigned to each fossil palynomorph throughout this paper are derived from Macphail et al. (1994), Macphail (1999) and Macphail et al. (2015). Briggs et al. (2007) published detailed carbon isotope data together with colourimetry data for the M1B seam based on two cores (LY4017, LY4018) taken from Loy Yang Open Cut (Fig. 2), with carbon isotope analyses at intervals of 0.1 m. The colour measurements of Briggs et al. (2007) were calculated using an alternate colour measurement technique (L*, a*, b*) and only the lightness variable (L*) was presented. While there is much variability in the lightness values of Briggs et al. (2007), it is possible to correlate our measured colour index log for the M1B seam by aligning the lightest peaks of L* (Briggs et al., 2007) with those determined by Holdgate et al. (2014). This correlation allows our colourimetry log to be matched with the carbon isotopic δ13C values of Briggs et al. (2007) to a reasonable degree of stratigraphic accuracy. Macro and micro charcoal were quantified to assess the distribution of charcoal in the lithotypes from the M1B and M2A seams of the Latrobe Valley. Microscopic charcoal particles, within the palynological slides, were point counted with the identification of charcoal based on it being black, uniformly opaque, brittle and angular (Waddington, 1969; Patterson et al., 1987; Mooney and Tinner, 2011). Macroscopic charcoal fragments were counted in two classes (250 m and 125 m), using the method of Clark (1982). Charcoal preparation was based on the method of Sander and Gee (1990). Charcoal residues were picked under a low power binocular microscope (Scott, 2010) in a Petri-dish. In addition, these samples were weighed before and after this process was undertaken, ensuring the amount of charcoal (wt.%) could be estimated. Scanning electron microscopy (SEM) of charcoal fragments was undertaken on a Philips FEI XL30 environmental SEM on gold-coated samples. Reflected light microscopy was carried out on polished blocks. Quantitative colourimetry was also completed on modern charcoal.

239

Fig. 3. Measured stratigraphic sections through the M1B and M2A seams at Loy Yang Open Cut illustrating the occurrence and abundance of charcoal in the laminated dark and dark lithotypes based on macro charcoal counts. The brown interval represents the distribution of palynology samples.

5. Results 5.1. The #64 m lightening upwards cycle A detailed palynological study was carried out in the upper portion of the M1B seam (stratigraphic heights from 65 to 85 m, Fig. 3). Holdgate et al. (1995) subdivided the M1B seam into subseams and

240

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

named the prominent dark horizons, by their stratigraphic height above the base of the seam (e.g. #24 m, #35 m, #64 m, #78 m). The sampled stratigraphic section for this study is derived from the southern side of Loy Yang Open Cut (Fig. 2) and yielded similar but not identical stratigraphic heights (Holdgate et al., 2014). As such, the #64 m cycle (defined by the presence of a thick, laminated dark lithotype unconformably overlying a lighter lithotype) is equivalent to the horizon measured at 67.1 m above the base of the seam in this study. The #64 m lightening upwards cycle samples are within the Upper Proteacidites tuberculatus Zone based on the presence of Psilastephanocolporites micus Partridge 1973, which first appears at the base of the subzone according to the biostratigraphy of Stover and Partridge (1973). The lack of indicators from the overlying younger zone reinforces the subzone assigned to these samples and suggests an Early Miocene age for the samples (Fig. 4). The studied interval has a complete lightening-upwards cycle and detailed descriptions of select key taxa and charcoal, within each lithotype, in the cycle are described below. 5.1.1. Interval 65 to 66 m – medium light lithotype The upper part of the previous shallowing-upwards cycle (from 65 to 66 m) consists of the medium light lithotype. This lithotype is characterised by high abundances of the Nothofagus subgenera Brassospora and Fuscospora, Casuarinidites cainozoicus (Casuarinaceae), Periporopollenites demarcates (Trimeniaceae) and Phyllocladidites mawsonii (Lagarostrobos franklinii type). It is also associated with moderate abundances of Quintiniapollis spp. (Quintinia), Sapotaceoidaepollenites spp. (Sapotaceae) and Elaeocarpaceae (Fig. 5b). The medium light lithotype had minor amounts of micro charcoal and no macro charcoal fragments recorded. This lithotype has a heavy carbon isotopic value of ~− 25.5‰ pdb (Fig. 5c) and at Loy Yang contained a low abundance of large wood fragments in outcrop. 5.1.2. Interval 67.1 to 68.5 m - laminated dark lithotype The lowest part of the cycle (from 67.1 to 68.5 m) consists of the laminated dark lithotype that is separated from the underlying medium

light lithotype by a prominent erosional surface. This lithotype is characterised by relatively high abundances of Aglaoreidia qualumis (Typhaceae), Cyperaceaepollis spp. (Cyperaceae) and Milfordia homeopunctata (Restionaceae). The abundance of the ferns Gleicheniidites spp. (Gleicheniaceae), Cyathidites spp. (Cyatheaceae/ Lindsaeaceae), Matonisporites ornamentalis (Dicksonia), Dictyophyllidites spp. (Gleicheniaceae), and Ericipites spp. (Ericaceae) is highest in this lithotype (Fig. 5a). The Nothofagus subgenera Brassospora and Fuscospora, Araucariacites australis (Araucariaceae) and Myrtaceidites spp. (Myrtaceae) are also minor components of this lithotype. The highest abundance of both macro and micro charcoal was recorded in the laminated dark lithotype in the M1B seam (Fig. 5c). A large portion of the micro charcoal was featureless while the macro charcoal consisted of large fragments N5 cm in length (Fig. 6). Charcoal constitutes up to 50% (by weight) of the laminated dark coal. Individual charcoal fragments under the SEM revealed well preserved, charcoalified stem fragments showing homogenised cell walls. Highly reflective, charcoalified stem fragments were also observed in polished blocks (Fig. 7). Based on the carbon isotope data of Briggs et al. (2007), the laminated dark lithotype has relatively uniform and heavy carbon isotope values (average around −25.75‰ pdb). A total absence of large wood fragments is observed in the laminated dark lithotype. 5.1.3. Interval 69 to 72 m – dark lithotype Overlying the laminated dark lithotype is an interval of unlaminated (massive) dark lithotype (69 to 72 m). This interval is characterised by a sharp increase in the abundance of gymnosperms (Fig. 5a), predominantly the Podocarpaceae Dacrycarpidites australiensis (Dacrycarpus) and Dacrydiumites florinii (Dacrydium). Microalatidites palaeogenicus (Phyllocladus), Microcachrydites antarcticus (Microcachrys tetragona), Phyllocladidites mawsonii (Lagarostrobos franklinii type) and Podocarpidites spp. (Podocarpus) also increase in abundance. Cyathidites spp. (Cyatheaceae/Lindsaeaceae), Elaeocarpaceae, Cunoniaceae (Weinmannia type), Proteacidites spp. (Proteaceae) and Casuarinidites cainozoicus (Casuarinaceae) are also found in relatively high

Fig. 4. Stratigraphy of the Oligocene-Miocene Gippsland Basin. Gradstein et al. (2012) timescale used. Planktonic foraminifera derived from Berggren et al. (1995), Wade et al. (2011) and Taylor (1966). Spore-pollen zonation derived from Partridge (2006) and Stover and Partridge (1973). Coal seam ages from Holdgate et al. (1995).

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

241

Fig. 5. A) Palynological counts for the interval of 65 to 85 m in the upper part of the Morwell 1B Seam. B) Palynological counts for the interval of 65 to 85 m in the upper part of the Morwell 1B Seam. C) δ13C%, micro charcoal and macro charcoal values from the interval 65 to 85 m in the upper part of the Morwell 1B Seam.

242

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

Fig. 5 (continued).

abundances (Fig. 5b). Ericipites spp. (Ericaceae), the ferns Gleicheniidites spp. (Gleicheniaceae), Dictyophyllidites spp. (Gleicheniaceae) and Matonisporites ornamentalis (Dicksonia) have a relatively low abundance in this lithotype. Aglaoreidia qualumis (Typhaceae), Cyperaceaepollis spp. (Cyperaceae) and Milfordia homeopunctata (Restionaceae) are completely absent. The Nothofagus subgenera Brassospora and Fuscospora and Araucariacites australis (Araucariaceae), despite increasing slightly, is a relatively minor component of the sporepollen assemblage recorded from this lithotype. Macro and micro charcoal are abundant in this interval (Fig. 5c). However, charcoal makes up a relatively smaller component of the coal than in the underlying laminated dark lithotype. Carbon isotope values for this interval display variable and light values. Wood makes up a minor component of this interval. 5.1.4. Interval 73 m to 82 m – medium dark lithotype The medium dark lithotype is characterised by an increased abundance of angiosperms, notably Anacardiaceae, Casuarinidites cainozoicus (Casuarinaceae), Elaeocarpaceae, Quintiniapollis spp. (Quintinia),

Sapotaceoidaepollenites spp. (Sapotaceae). The Nothofagus subgenera Brassospora and Fuscospora is a moderate component of the spore-pollen recorded from this lithotype, increasing in abundance upwards in the cycle (Fig. 5b). Significantly, Nothofagidites deminutus (Nothofagus subgenus Brassospora) shows a dramatic increase in abundance at about 74 m, from being a very minor component below this level, to of great abundance. Gleicheniidites spp. (Gleicheniaceae), Dictyophyllidites spp. (Gleicheniaceae), Matonisporites ornamentalis (Dicksonia) and Cyathidites spp. (Cyatheaceae/Lindsaeaceae) are not recorded in this lithotype. The medium dark lithotype has minor amounts of micro charcoal recorded and no macro charcoal fragments (Fig. 5c). The medium dark lithotype has a relatively lighter carbon isotope composition (~ − 26.25‰ pdb) and contained an increased abundance of large wood fragments. 5.1.5. Interval 83 to 83.5 m - laminated dark and dark lithotypes The lowest part of the next cycle (83 to 83.5 m) consists of a laminated dark to dark lithotype. An unconformity separates the laminated

Fig. 6. A–B) Photograph of micro-charcoal (arrows) in a palynology slide from the laminated dark lithotype (depth 67.1 m) of the Latrobe Valley brown coals. C) Laminations observed at the #64 m cycle boundary exposed at Loy Yang Open Cut. D) Charcoalified plant fragments from the laminated dark lithotype exposed at Loy Yang Open Cut (depth 67.1 m).

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

243

Fig. 7. Examples of charcoal specimens found in the laminated dark lithotype from the Latrobe Valley. A) Scanning electron micrograph of Gymnostoma (Mike Pole, pers. comm.). B) Scanning electron micrograph of well preserved, charcoalified stem fragments showing homogenised cell walls. C–D) Highly reflective, charcoalified stem fragments in polished blocks.

dark lithotype from the underlying cycle. This lithotype is characterised by relatively high abundances of Gleicheniidites spp. (Gleicheniaceae), Cyathidites spp. (Cyatheaceae/Lindsaeaceae), Matonisporites ornamentalis (Dicksonia), Dictyophyllidites spp. (Gleicheniaceae), Milfordia homeopunctata (Restionaceae) and Ericipites spp. (Ericaceae) (Fig. 5a). The Nothofagus subgenera Brassospora and Fuscospora and Myrtaceidites spp. (Myrtaceae) are also minor components of this lithotype. Both macro and micro charcoal were recorded from this interval (Fig. 5c). From the carbon isotope data of Briggs et al. (2007), this interval has relatively heavy carbon isotope values (average around − 25.25‰ pdb). A total absence of large wood fragments is observed in this interval. 5.1.6. Interval 84 to 85 m - medium dark lithotype This medium dark lithotype is characterised by an initial increase in the abundance of gymnosperms, specifically the Podocarpaceae, followed by relatively high abundances of Casuarinidites cainozoicus (Casuarinaceae), Quintiniapollis spp. (Quintinia), Elaeocarpaceae, Proteacidites spp. (Proteaceae) and the Nothofagus subgenera Brassospora and Fuscospora (Fig. 5b). Ericipites spp. (Ericaceae), the ferns Gleicheniidites spp. (Gleicheniaceae), Dictyophyllidites spp. (Gleicheniaceae) and Matonisporites ornamentalis (Dicksonia) have a relatively low abundance in this lithotype (Fig. 5a). Macro and micro charcoal are still relatively abundant in this interval. However, charcoal makes up a much smaller component of the coal than in the underlying laminated dark to dark lithotype (Fig. 5c). Carbon isotope values for this interval display very light values (average around − 26.25‰ pdb). Wood makes up a minor component of this interval.

carbon isotope values (Figs. 8–9). The laminated dark lithotypes in the M1B seam have heavy carbon isotope values, which average around − 25.25‰ pdb. The dark lithotypes have heavy carbon isotope values (average around −25.5). The medium dark lithotypes throughout the M1B seam have relatively lighter carbon isotope compositions (average around −26.25‰ pdb), and the medium light lithotypes have light carbon isotope compositions (average around − 26‰ pdb). The light lithotypes in the M1B seam have heavy carbon isotope values (average around − 24.5‰ pdb) while the pale lithotypes have exceptionally heavy carbon isotope values (average around −23.75‰ pdb). The colourimetry results from the M1B seam revealed a strong correlation between yellow and red colour pigments (Fig. 10). Throughout the M1B seam, the laminated dark lithotype has very low ‘a’ (red) and ‘b’ (yellow) values on the scatter plot of colour. This lithotype is the only lithotype to extend into negative ‘a’ values, reflecting the colour blue. The dark lithotype has low ‘a’ (red) and ‘b’ (yellow) values on

5.2. General characteristics of the M1B seam lithotypes Throughout the M1B seam charcoal was most abundant and recorded throughout all but one of the laminated dark lithotypes (Fig. 3). Both macro and micro charcoal were recorded in the charcoal-bearing laminated dark lithotypes in the M1B seam, with varying amounts of charcoal recorded in each of the laminated dark lithotype intervals. The M2A seam also contained minor macro charcoal in dark lithotypes. The lithotypes throughout the M1B seam, based on an averaged colour index value for each lithotype, display a correlation with median

Fig. 8. Correlated section between L*, δ13C and colour index (CI) at Loy Yang Open Cut. L* and δ13C data from Briggs et al. (2007) and colour index data from Holdgate et al. (2014).

244

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

Fig. 9. Correlation between lithotype and δ13C based on average colour index values from Holdgate et al. (2014) and median δ13C values from Briggs et al. (2007).

the scatter plot of colour, which increase slightly into the medium dark lithotype. The medium light lithotype has higher ‘a’ (red) and ‘b’ (yellow) values on the scatter plot of colour as do the light lithotypes. The pale lithotypes have exceptionally high ‘a’ (red) and ‘b’ (yellow) values on the scatter plot of colour. In contrast, charcoal (modern) has exceptionally low or negative ‘a’ (red) and ‘b’ (yellow) values. 6. Discussion 6.1. Depositional setting of lithotypes in the #64 m cycle There has been considerable debate about the depositional environment of lithotypes in the Australian and German brown coals with contrasting models, termed the dry-dark and dry-light models, being previously proposed for the origin of lithotypes (Holdgate et al., 2014; Holdgate et al., in press). The dry-dark model proposed that darker lithotypes were produced in the driest facies, while light lithotypes were deposited in either very wet or open water environments (Teichmüller, 1958; Kershaw and Sluiter, 1982; Teichmüller, 1989), based on a predominance of pollen derived from dryland plants with good dispersal ability in the lighter lithotypes, a lack of identifiable macroremains (Luly et al., 1980; Sluiter, 1984; Kershaw et al., 1991), and organic geochemistry (Finotello and Johns, 1986). An ‘open water’ depositional environment for the lighter lithotypes based on the

Fig. 10. Scatter plot of red versus yellow colour pigments used to derive the colour index values from the M1B Seam determined for the various lithotypes.

palynology alone was criticized by Anderson and Mackay (1990) who argued that a lack of pollen from the ‘local vegetation’ did not necessarily imply ‘open water’ conditions. Moreover, the lack of macroscopic plant debris in the pale and light coals was not consistent with deposition in ‘open water’ environments as plant debris commonly falls into lakes, conditions generally leading to the preservation of macroscopic structure (Anderson and Mackay, 1990). Instead this implied that conditions at the time of deposition actually led to high degrees of degradation (i.e. drier environments for the lighter lithotypes) (Anderson and Mackay, 1990), a possibility not dismissed by Finotello and Johns (1986) who stated that more work was needed to draw firmer conclusions. The dry-light model contrastingly proposed that light lithotypes were deposited in the driest facies, while dark lithotypes were deposited in the wettest facies (Hagemann and Hollerbach, 1980; Hagemann and Wolf, 1987; Anderson and Mackay, 1990; Holdgate et al., 2014), based on multidisciplinary studies that involved using the geochemical composition, colour, palaeobotany, palynology and coal petrology. This controversy is even more complex because lithotypes are arranged in cycles, which are generally attributed to being caused by terrestrialization (drying-upwards cycles) (e.g. Muller et al., 2003; Frank and Bend, 2004). This would imply that in the dry-dark model, cycles should darken upwards, whereas in the dry-light model, cycles should lighten upwards. In the Latrobe Valley lignites, Mackay et al. (1985) showed statistically (Markov Chain analysis) that the majority of cycles lighten upwards. These lightening-upwards cycles are obvious in weathered mine faces, and also on quantitative colourimetry logs (e.g. Fig. 3). 6.1.1. Laminated dark lithotype The high abundances of bulrush (Typhaceae), the common sedge (Cyperaceae) and cord-rush (Restionaceae) pollen, fossil spores of Gleicheniidites spp. and Dictyophyllidites spp. from the coral-fern family Gleicheniaceae, Cyathidites spp. (Cyatheaceae/Lindsaeaceae), Matonisporites ornamentalis (Dicksonia) and the heaths Ericipites spp. (Ericaceae) in the laminated dark lithotype suggest deposition in an emergent marsh or meadow marsh environment. In particular, the presence of the coral-fern family Gleicheniaceae indicates an open well-lit environment and suggests that the facies was not forested (e.g. Jones and Clemesha, 1976; McGlone and Neall, 1994). The presence of Typhaceae (entirely restricted to this lithotype) suggests subaqueous conditions. This palynological assemblage is also interpreted to have poor dispersal capacity (Kershaw et al., 1991), suggesting that this assemblage is derived from vegetation of local origin (rather than being transported). The highest abundances of micro and macro charcoal were recorded from the laminated dark lithotype. The large size of some charcoalified plant fragments, including sedges, rushes and delicate pinnules from

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

the fern Gleichenia (Blackburn and Sluiter, 1994) suggests that the fires were of low intensity. The preservation of these delicate and well-preserved fragments is dependent on the temperature of the original charcoalification (Vaughan and Nichols, 1995) with plant fragments becoming increasingly brittle as temperature increases (Scott, 2010). It is proposed that these charcoalified plant fragments fell into the water in and around the adjoining swamps. This is consistent with the observations made by Holdgate et al. (2014), that the lamination in the facies is indicative of subaqueous deposition. Recent studies on charcoal indicate that large fragments of charcoal (N125 μm) are derived in situ or transported very short distances and produced by local fires (Collinson et al., 2007; Nichols et al., 2000; Perry et al., 2014). A high proportion of the charcoal particles were larger than 250 μm, suggesting local fire activity at the time of deposition. The laminated dark lithotype is therefore interpreted to represent an emergent marsh with bulrushes that transitions into a meadow marsh with reeds, heaths and coral ferns (Fig. 11). 6.1.2. Dark lithotype The abundance of gymnosperm pollen, (specifically Podocarpaceae) including Dacrycarpidites australiensis (Dacrycarpus), Dacrydiumites florinii (Dacrydium), Microalatidites palaeogenicus (Phyllocladus), Microcachrydites antarcticus (Microcachrys tetragona), Phyllocladidites mawsonii (Lagarostrobos franklinii type) and Podocarpidites spp. (Podocarpus) increases sharply in the dark lithotype in the cycle examined. Araucariacites australis (Araucariaceae), Elaeocarpaceae, Cunoniaceae (Weinmannia type), Proteacidites spp. (Proteaceae) and Casuarinidites cainozoicus (Casuarinaceae) pollen are also relatively abundant in the dark lithotype. The abundance of these arborescent species indicates the transition from a meadow marsh to a forested bog environment upwards in this facies. The decline in the abundance of Ericipites spp. (Ericaceae), the ferns Gleicheniidites spp. (Gleicheniaceae), Dictyophyllidites spp. (Gleicheniaceae) and Matonisporites ornamentalis (Dicksonia) is consistent with lower light levels at the ground surface, consistent with the transition to a forested environment. The absence of lamination in this and the overlying facies suggests that the depositional environment was not subaqueous. Macro and micro charcoal are present in relatively low abundance within the

245

dark lithotype, indicating a significant reduction in fire frequency within this facies relative to the underlying laminated dark lithotype. The dark lithotype is interpreted to represent a transition from meadow marsh to forested bog, with Podocarpaceae being the dominant trees (in particular Dacrycarpus and Dacrydium). The basal portion of this lithotype likely represents the transition from a meadow marsh to a forested bog environment, while the upper portion of the lithotype is interpreted as a forested bog setting (Fig. 11). 6.1.3. Medium dark and medium light lithotypes These lithotypes have an increased abundance of angiosperms, including Anacardiaceae, Cunoniaceae, Elaeocarpaceae, Quintiniapollis spp. (Quintinia), Periporopollenites demarcates (Trimeniaceae), Proteacidites spp. (Proteaceae), Sapotaceae and the Nothofagus subgenera Brassospora and Fuscospora that proliferated in the Gippsland Basin during the Early Miocene due to the warm and wet climate (Gallagher et al., 2001). Phyllocladidites mawsonii (Lagarostrobus franklinii type) is also abundant in this lithotype, whereas there is a marked decline in ferns. The abundance of these arborescent taxa (and absence of ground cover taxa like Gleicheniaceae) suggests that the facies was forested and dominated by angiosperms (rather than gymnosperms as occur in the dark lithotype). The medium lithotypes have extremely minor micro charcoal and no macro charcoal recorded, indicating very low fire frequencies, consistent with deposition in a forested bog. The medium lithotypes are interpreted to represent an ombrogenous forest bog with angiosperms being the dominant tree (predominantly the Nothofagus subgenus Brassospora). The basal portion of the medium dark lithotype also represents the transition from a forested bog, dominated by podocarps to an ombrogenous angiosperm dominated forest bog (Fig. 11). 6.2. Allochthonous vs autochthonous palynological taxa A major concern in palynological studies is the degree of spore-pollen transport, both in air and water. The presence of taxa in a palynological assemblage with long distance dispersal might therefore easily be misinterpreted as autochthonous to that facies. Previous interpretations

Fig. 11. Facies model of the vegetation succession from an emergent marsh to a forested ombrogenous bog in the Early Miocene peatlands of the Latrobe Valley. Lithotype, substrate wetness and the degradation of the substrate are illustrated.

246

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

of the palynology from the Latrobe Valley brown coals have classified the palynomorphs into regional and local genera, depending on their good or poor pollen dispersal capabilities (e.g. Luly et al., 1980; Kershaw et al., 1991). This classification is generally based on studies of modern genera and their observed pollen dispersal capabilities (see Luly et al., 1980 for a review of Australian taxa). Taxa like Nothofagus and Quintinia have been noted as having good pollen dispersal (via wind) and have therefore been previously interpreted as being of regional, rather than local, origin. Nothofagus in particular, has been interpreted in this manner with the common assumption, based on previous Latrobe Valley studies, that Nothofagus forests dominated the surrounding mountainous regions (e.g. Luly et al., 1980; Kershaw et al., 1991) and were not growing within the peatlands themselves. Kershaw et al. (1991) listed the subgenera types (brassii, fusca and menziesii) of Nothofagus as being non-local, regional taxa. The lighter lithotypes, which contain an abundance of Nothofagidites pollen and an absence of Nothofagus macrofossils, were subsequently interpreted as subaqueous, with all palynomorphs being regionally transported. However, the use of modern taxa and pollen dispersal capabilities does have its limitations. Modern taxa, for example, may not be appropriate analogues for ancient species. Perhaps more importantly, the fact that a particular species has good pollen dispersal does not guarantee that this species is not autochthonous. Another approach that can give important information about the origin of palynomorphs is facies control. Palynomorph taxa that are restricted to a particular facies are likely to be of autochthonous, rather than allochthonous origin (if the taxa were transported, they should be present to some extent in all facies). Examples of strongly facies-controlled palynomorphs from the laminated dark lithotype in the #64 m cycle are the Restionaceae, Ericaceae and Gleicheniaceae. These taxa are abundant in the laminated dark lithotype but are essentially absent from other lithotypes, suggesting an autochthonous origin. Consistent with this observation, Luly et al. (1980) and Kershaw et al. (1991) classified these taxa as having poor pollen dispersal capabilities. In the dark lithotype, the Podocarpaceae are abundant, but again are almost completely absent from the other lithotypes (particularly the laminated dark lithotype), suggesting an autochthonous origin. Consistent with this, most of the Podocarpaceae are suggested to have poor pollen dispersal and are derived from local or extra-local sources (Luly et al., 1980; Kershaw et al., 1991). In contrast, taxa like Casuarinaceae and Nothofagus (total counts) are interpreted as regional flora (Kershaw et al., 1991). These taxa are present in all lithotypes, with a general increase in the lighter lithotypes. However, the distribution of one particular species, Nothofagidites deminutus Cookson 1959 (Nothofagus subgenus Brassospora) appears to be strongly controlled by facies. Nothofagidites deminutus (Nothofagus subgenus Brassospora) is most abundant in the lighter lithotypes and almost completely absent in the other lithotypes, suggesting an autochthonous origin for this species. The recent discovery of abundant Nothofagus subgenus Brassospora leaves by Carpenter et al. (2014) within Oligo-Miocene lignites in New Zealand also suggests that the Nothofagus subgenus Brassospora could and did grow in lowland, ombrotrophic mires, again implying an autochthonous origin for this Nothofagus species in a peatland setting. Similarly Quintiniapollis spp. (Quintinia) is strongly facies controlled, being restricted to medium and light lithotypes. Quintinia was suggested to be a regional taxon by Kershaw et al. (1991), despite its strong facies restriction. It is significant that all of the lithotypes described here have a floral component that is facies-controlled. This suggests that all of the lithotypes have an autochthonous floral component. 6.3. Lithotype-floral correlations for the Latrobe Valley brown coals The detailed palynological profile from the #64 m lithotype cycle within the M1B seam at Loy Yang demonstrates that lithotype and quantitative colourimetry have a strong correlation with floral data.

New palynological and charcoal data indicate that the laminated dark lithotype represents a fire-prone emergent marsh with bulrushes that transitions into a meadow marsh with reeds, heaths and coral ferns. This interpretation is consistent with the findings of Holdgate et al. (2014) that the laminated dark lithotype represents an open water/submerged stage, based on the laminated nature of the facies. The interpretation of the dark lithotype representing the transition from a meadow marsh to a podocarp-dominated forested bog is based on new detailed palynological data and well constrained colour index values (i.e. lithotype assignment). This interpretation modifies the findings of Holdgate et al. (2014) who proposed that the dark lithotype likely represents a rush/heath moor stage, based on a lack of wood, an abundance of rush cuticle and rush, heath and fern pollen and spores. This modification largely results from the more detailed stratigraphic sampling used in this study (25 cm spacing) relative to that of Sluiter's (1984) original palynology study (N 1 m average sample spacing) upon which Holdgate et al. (2014) based their results. Moreover, the lithotype subdivision of Sluiter (1984) did not include the laminated dark lithotype, resulting in this facies being grouped into the dark lithotype. Furthermore, the colourimetry measurement of the present study has a higher precision that that of Sluiter's (1984) study (probably due to improved analytical precision of the modern Chromameter). The medium lithotypes are interpreted to represent the transition from a forested bog, dominated by podocarps to an ombrogenous forest bog with angiosperms being the dominant tree (predominantly the Nothofagus subgenus Brassospora). This is somewhat consistent with the findings of Holdgate et al. (2014) that the medium dark lithotypes may correspond to a marginal woodland stage (as indicated by an increased abundance of large wood fragments), while the medium light lithotypes corresponded to an ombrogenous forest based on the presence of arborescent taxa. The model presented here is consistent with previous palynological studies undertaken. Luly et al. (1980) carried out a detailed palynological study of the Yallourn seam, both in core, and in outcrop while Sluiter (1984) undertook a detailed palynological study of the Morwell 1B seam from core. Both Luly et al. (1980) and Sluiter (1984) measured colour on their samples, so general comparisons can be made between our results and theirs. Despite these authors not recognizing lightening-upwards cycles and not recognizing the laminated dark lithotype, the palynological results of these authors show many similarities with our study. Both Luly et al. (1980) and Sluiter (1984) found an abundance of Restionaceae and Gleicheniaceae in the darkest lithotypes, and abundant podocarps in the dark and medium-dark lithotypes. The Yallourn Seam study by Luly et al. (1980) also revealed high abundances of angiosperms including Nothofagus, Casuarinaceae, Quintinia and Sapotaceae in the lighter lithotypes. Blackburn (1981, 1985) concurrently undertook a detailed study of macrofossils recovered from Yallourn Open Cut and concluded that coal lithotypes are strongly related to the fossil floras contained in them. Blackburn (1985) found an abundance of Gleicheniaceae foliage/roots, Typhaceae and Cyperaceae leaves in the darkest lithotype. Blackburn (1981) noted high proportions of fern and conifer fossils, assigned to Dacrydium and Callitris (Cupressaceae), associated with reduced angiosperm diversity in the dark lithotypes. Blackburn (1981) also reported that angiosperm leaves and cuticles were increasingly abundant and diverse in the medium-light lithotypes while podocarps and conifers were very rare. The lightest coals were found to contain small fragments of angiosperm leaves, rare and fragmented angiospermous cuticles and very rare podocarp leaves (Blackburn, 1981). In summary, Blackburn (1981, 1985) found from macrofossils that ferns and podocarps were most prevalent in the darker lithotypes while angiosperms became increasingly abundant in the lighter lithotypes. This is consistent with palynological data from this study and also suggests that much of the palynoflora in the coals is autochthonous.

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

6.4. Modern ecological analogues for the Latrobe Valley peatlands The extensive coastal, peats of the Indo-Malayan archipelago have been considered the best modern sedimentological analogue for the Latrobe Valley peatlands (Holdgate et al., 2014) based on their size and lightening upwards succession, resulting from peat aggrading and drying above the water table (ombrogenous domes) (Esterle and Ferm, 1994; Cameron et al., 1989). However, it is clear that the Indonesian peatlands are poor floral analogues for the brown coal facies of the Latrobe Valley, with very few taxa being present in both settings. In contrast, a significant proportion of the fossil taxa recorded in this study have affinities with plant communities existing in the lowlands of southern New Zealand. New Zealand's modern flora, which includes extensive Nothofagus forests (Ogden and Stewart, 1995; Ogden et al., 1996), has predominantly Austral affinities perhaps because a ‘gene flow’ was maintained between Australia and New Zealand following their separation by long-distance dispersal of pollen, spores and seeds (Mildenhall, 1980). Common genera in the M1B seam occur within modern New Zealand forests and grow in environments similar to those interpreted for the M1B lithotypes. Examples include Dacrycarpus, Dacrydium, Podocarpus, Cunoniaceae (i.e. Weinmannia) and Nothofagus (subgenus menziesii), which dominate the lowland forests of New Zealand today (Wilmshurst et al., 2002) and also occur in the Latrobe Valley coals. These same genera also dominated the ombrogenous forest mires that formed the extensive peatlands of the Gore Lignite Measures in New Zealand (Carpenter et al., 2010; Ferguson et al., 2010). The similarity, not only in the floral taxa present, but in the vegetation successions makes the well-studied modern New Zealand flora (in particular, South Westland, the Waikato Region and Northland) a good floral analogue for the Latrobe Valley peatlands. In general in the New Zealand swamps, an initial marsh phase dominated by sedges (Cyperaceae), ferns (Gleicheniaceace), Typhaceae, Restionaceae and Leptospermum (Wardle, 1974; Mark and Smith, 1975; Mew, 1983; Newnham, 1992; Clarkson et al., 2004; Deng et al., 2006a) characterises the initial phase of wetland development. This floral association is very similar to that found in the laminated dark and lower portion of the dark lithotype in the M1B Seam. This marsh phase in New Zealand is commonly overgrown by a tall podocarp forest, dominantly consisting of Dacrydium stands (Holloway, 1954; Wilmshurst et al., 2002). Again this is very similar to the upper portion of the dark lithotype in the M1B Seam. In New Zealand, the podocarp genera, Dacrycarpus and Dacrydium, are most suited to growing in silty peats (Wardle, 1974), bogs, swamps (Hunt, 2007) and wetter substrates (Wilmshurst et al., 2002). Leathwick and Austin (2001) found that these conifers can readily tolerate wet substrates largely at the expense of Nothofagus, which decline in moist and fertile areas with poor drainage (Wardle, 1964). Other studies undertaken in New Zealand show that dense Podocarp forests are common on the poorly-drained, flat lowlands (Leathwick, 1995; Wardle, 1980) where Dacrydium and Dacrycarpus both reach their highest densities (Wardle, 1980; Stewart et al., 1993; Leathwick and Austin, 2001), potentially a result of their extensive surface/subsurface lateral roots (Norton, 1989; Hunt, 2007). Other plant communities that can thrive in poorly-drained environments in New Zealand consist of Weinmannia, Quintinia, Elaeocarpaceae, Phyllocladus, Gleicheniaceae and the tree ferns (Dicksonia and Cyatheaceae) (Mark and Smith, 1975; Wardle, 1977; Wilmshurst et al., 2002), all genera that occur within the OligoMiocene peatlands of the Latrobe Valley. As in the M1B lithotype cycles, the overgrowth of Podocarp forests by predominantly angiosperm forest species is also documented in New Zealand. Newnham (1992) and Holloway (1954) noted a decrease in the abundance of gymnosperm taxa in vegetation successions, and their subsequent replacement with angiosperms, including Nothofagus and Quintinia. Robbins (1962) also observed that vegetation successions in New Zealand terminate with the dominance and rise of the shade-

247

tolerant Nothofagus due to the gradual replacement of the light demanding podocarps in the mixed podocarp-broadleaved forest. Wilmshurst et al. (2002) documented the occurrence of species in relation to substrate wetness within a modern New Zealand peat bog (on the South Island). Surface wetness in this peat study was assessed by analysis of testate amoebae, peat humification and plant macrofossils. They found that Dacrycarpus and Dacrydium are associated with the wettest environments. As the substrate dries, Podocarpus dominates, followed by Weinmannia and finally Nothofagus. This order of genera occurrence, from the wettest to driest environment, precisely mimics that found in the #64 m lightening-upwards cycle from base to top. This similarity between a New Zealand peat bog and a lightening-upwards cycle from the Latrobe Valley coals suggests: 1. New Zealand's vegetation communities may represent a modern floral analogue for vegetation successions in the Oligo-Miocene coals of the Latrobe Valley. 2. Lightening-upwards cycles from the Latrobe Valley coals probably represent drying-upwards environments. The similarities between the modern floral associations of New Zealand and the Oligo-Miocene floras of the Latrobe Valley can be explained by the evolutionarily conservative nature of the principal taxa involved. In the case of podocarps, many Eocene and Early Oligocene species are almost identical to extant species (e.g. the fossil Microstrobos microfolius versus the extant M. niphophilus, Brodribb and Hill, 2004). Brodribb and Hill (2004) also noted that Cenozoic floral associations are similar to modern associations, giving the example of podocarps being associated with Weinmannia, an association that also occurs in the brown coals of the Latrobe Valley. 6.5. Environment vs climate in cycle development As discussed above, the vertical palynological succession described here from the Oligo-Miocene M1B seam is very similar to that described from many Holocene peat successions in New Zealand. In general, a marsh flora rich in sedges, ferns and restiads represents the initiation of peat accumulation and the base of the M1B cycles and Holocene New Zealand peats. Overlying is a flora with abundant podocarps (particularly Dacrycarpus and Dacrydium), Weinmania and Elaeocarpaceae. The uppermost floral facies has a greater abundance of angiosperms, notably Quintinia, and Nothofagus, although some podocarps are also present. Excluding species that are endemic to either New Zealand or Australia (e.g. Phormium in New Zealand and Casuarinaceae in Australia), the flora is near identical. Given that the flora within individual facies and within the succession as a whole are remarkably similar, a common control is indicated. The question then is “what is this control?” Two very different interpretations can, and have been made regarding floral successions in peatlands. 1. The successions largely record climate change (allogenic control); and 2. The successions record local environmental change in the peatland (autogenic control). In the climate change hypothesis, palynomorphs are largely interpreted as allochthonous to the peatland and wetland development is controlled by climate (e.g. McGlone and Bathgate, 1983; McGlone, 1988; Wilmshurst et al., 2002; McGlone, 2009). In contrast, peatland floral successions are commonly interpreted as being controlled by wetland evolution (hydroseral succession) related to filling of the initial water body and growth of the peat above the water table (e.g. Anderson, 1964; Anderson and Muller, 1975; Morley, 1981; Muller et al., 2003). Many peatland researchers have considered the influence of the local environment and wetland succession on vertical peatland sections (e.g. Rydin and Jeglum, 2006; Rochefort et al., 2012). Bunting et al. (2008) concisely documents problems in the interpretation of peat stratigraphy that relate to wetland system evolution. Bunting et al. (2008) suggested that it is difficult to separate the regional dryland signal from the autochthonous component of the wetland and that peats

248

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

cannot be treated as simple collection points for the regional dryland flora. Moore and Webb (1978) and Traverse (1988) also established that the palynomorphs in peats and coals are predominantly of local derivation because the peat-producing flora contributes a large local pollen input (Moore et al., 1991; Williams et al., 1993). Traverse (1988) further concluded that the palynomorphs in coals represent a particular facies that tends to recur mostly in response to changes in the environment. Recent works that attribute floral variations in peat profiles to changes in climate (Wilmshurst et al., 2002; Li et al., 2008; McGlone, 2009) also highlight that other factors (i.e. the nature of the substrate or landscape change) contributed to the floral composition of swamps/peats (eg. Morley, 1981; McGlone, 2009). We suggest that the simplest explanation for the dominantly dryingupwards cycles in the Cenozoic brown coals is hydroseral succession (terrestrialization) within the wetland depositional setting. Walther's Law (see Walther, 1894) readily explains the remarkable similarity between numerous observed lateral floral zonations in modern New Zealand wetlands (eg. Wardle, 1964, 1974) and the succession observed in vertical profiles from Holocene peats and Cenozoic brown coals. There is little doubt that climate has played a role in peatland floral successions. However, as noted by Bunting et al. (2008), the climate signal from vertical peatland profiles must be carefully separated from the autogenic processes that invariably occur within wetland systems. The almost automatic assumption of a dry land/regional and transported origin for most palynomorphs in some Holocene peat studies is not justified by observational data on the way peatlands behave (eg. Cameron et al., 1989; Esterle and Ferm, 1994; Rydin and Jeglum, 2006; Rochefort et al., 2012). Better understanding the complex ecological interactions between species and their environment (Birks, 1981) will likely enable more reliable reconstructions of past climate from palynology data. 6.6. Fire The charcoal-rich laminated dark lithotype contains an array of taxa that in modern environments are fire-tolerant as first proposed by Holdgate et al. (2014). Restionaceae have fire-resistant belowground stems, and are able to rapidly re-sprout after fire, when there is a high proportion of bare ground (Pate et al., 1999; Johnson, 2005). Modern day Gleichenia are also reported to be fire-tolerant (Collinson, 2002), implying that the high abundances of Gleichenia spores and charcoalified pinnules of Gleichenia that resemble Gleichenia dicarpa is a reflection of its ability to tolerate fire (Blackburn and Sluiter, 1994). The co-occurrence of charcoal and high abundances of fern spores has also been observed by Riegel (2000) in palynological preparations from the Lower Eocene lignites and clastic interbeds of Helmstedt, Germany, suggesting that ferns were common elements in many fire-prone marsh environments, because of their preference to occupy wet and open areas (Jones and Clemesha, 1976). Many recent studies illustrate that terrestrial plants in marshes/ peatlands/wetlands, despite their water-saturated substrates, will burn, because these plant communities are exceedingly flammable (McGlone, 2009). The Typha, sedge and rush components of emergent and meadow marshes in New Zealand can become very dry after even short periods without precipitation, and provide a suitable medium in which fire can develop and spread (Norton, 1989; Rogers et al., 2007) after starting by lightning (Bond et al., 2004). Typha, found in floating marshes, but also rooted in sediments, is a highly productive plant (Conner et al., 1986; McK. Pegman and Ogden, 2005) and adds large amounts of dry and flammable organic matter to swamp surfaces annually (Deng et al., 2006b). Studies completed in New Zealand show that large raised bogs and lowland wetlands not only have a long history of natural fire (McGlone, 2009) but are also subject to repeated natural fires (Perry et al., 2014). These sites have physical conditions conducive to fire: they are raised, have a low and uniform vegetation cover (i.e. no natural firebreaks), endure a relatively dry period, and the restiad cover

generates a dry and heated surface (Campbell and Williamson, 1997). Post-fire successions in New Zealand are also often associated with either Typha, an invader of disturbed wetlands and swamps (Birks et al., 1976; Jackson et al., 1988; Deng et al., 2006b) or Gleichenia, favoured in post-fire succession unless the fire leads to a significant rise in the water table due to forest destruction (Crawford, 2014), favouring Typha. Again, New Zealand represents a good floral analogue for the charcoal-rich, laminated dark lithotypes of Latrobe Valley peatlands. Other fire-adapted taxa including Leptospermum, Myrtaceae and the tree fern Dicksonia, also occur in New Zealand and in the Latrobe Valley peatlands (Luly et al., 1980; Sluiter, 1984; Perry et al., 2014). It is therefore feasible that the Latrobe Valley peatlands were subject to repeated fires during the Late Oligocene to Middle Miocene. The flammable vegetation occupying the emergent and meadow marshes were likely to have periodically burnt, resulting in the deposition of charcoal in a subaqueous environment, that ensured its preservation in the laminated dark lithotype. The widespread areal coverage of the Latrobe Valley wetlands would also result in a high likelihood of lightning-induced wildfires. 6.7. Lithotype formation and chemistry Chemical studies undertaken on brown coals, in both Australia and Germany, demonstrate that various degradation processes contribute to the formation of the specific lithotypes discussed in this paper. Hagemann and Wolf (1987) and Winkler (1986) found that hopane concentrations were high in dark lithotypes from the Helmstedt deposits of West Germany, and from this inferred higher levels of bacterial activity in an inundated depositional environment. The degree of biochemical gelification, characterized by the infilling of cell and other tissue cavities by gelified organic matter within the interstitial waters, which at some stage has been in a “plastic phase”, is most abundant in the dark lithotypes of the Latrobe Valley (Anderson and Mackay, 1990). Likewise, Diessel (1992) interpreted the occurrence of gelification as indicative of deposition in a wet environment. The darker lithotypes were also found to have undergone more intense anaerobic decomposition (Blackburn, 1985; Verheyen et al., 1984; Finotello and Johns, 1986; Hagemann and Wolf, 1987). The medium dark lithotypes from the Morwell 1 seam in the Latrobe Valley had significant amounts of 2,6,10,15,19-pentamethyleicosane, a C25 isoprenoid hydrocarbon, regarded as a marker for methanogenic bacteria (Anderson and Mackay, 1990). This implies that the diagenesis of the biomass input into the medium dark lithotype took place under anaerobic conditions. The concentration of porphyrins in the medium dark lithotype is at least 1000 times greater than in the pale lithotypes (Anderson and Mackay, 1990). Porphyrins, especially chlorophyll, are ubiquitous inputs into developing peats, however, they are readily degraded under oxic conditions (Baker and Louda, 1986). Hence the higher porphyrin content of the medium dark lithotype indicates minimal aerobic degradation of the original plant material. The degradative pathways of a number of biomarkers observed in paler lithotypes are characterized by successive oxidation (Anderson and Mackay, 1990). Finotello and Johns (1986) reported that their paler sample consisted of 60–70% humic acids, which were highly aromatic, indicating aerobic degradation of woody tissue in the original peat (Anderson and Mackay, 1990). The degradation of lignin to a variety of simple aromatic and phenolic compounds by aerobic decomposition, and the incorporation of these into humic acid-like materials, is well documented (Martin et al., 1980). The total liptinite concentration also shows a general increase from the dark to light lithotypes. Liptinite is, for the most part, produced from aliphatic plant biopolymers, such as cutin, suberin and sporopollenin, which tend by their nature to be resistant to degradation, especially oxidative degradation (Anderson and Mackay, 1990). Humic acid and pollen increase in concentration from the dark to medium light lithotypes (Anderson and Mackay, 1990), suggesting increased degradation, due to oxidation or mechanical damage

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

(Twiddle et al., 2012), for the lighter lithotypes at the time of deposition. The lighter coals are reported to result from aerobic degradation of the peat deposits, when accumulation, or lowering of the prevailing water table, resulted in drying of the peat surface in contact with the atmosphere (Hagemann and Hollerbach, 1980; Hagemann and Wolf, 1987; Mackay et al., 1985; Anderson and Mackay, 1990). It is therefore suggested that the darker lithotypes represent deposition in an anoxic depositional environment that promoted anaerobic decomposition of the inundated or periodically inundated substrate while the onset of aerobic degradation is preserved in the medium light lithotype, deposited in a raised ombrogenous forest bog. These findings suggest that the colour index values reflect the depositional environment and associated degradation of the initial peat depending on either oxic (the lighter lithotypes) or anoxic conditions (the darker lithotypes). The light and pale lithotypes appear to result from major oxidation and aerobic degradation.

6.8. Carbon isotopic composition The detailed colourimetry and carbon isotope profile from the M1B seam at Loy Yang demonstrates that lithotype and quantitative colourimetry have a strong correlation with carbon isotopes. Previous studies (see Farquhar et al., 1989; Cloern et al., 2002) have found that variability in the carbon isotopic composition of plants can result from a number of factors including differences in the vegetation's photosynthetic pathway, the source of carbon dioxide and the plants water-use efficiency. Other factors can also influence the carbon isotopic δ13C values of peats and coals, resulting in different isotopic values for each individual lithotype. Charcoal and charred wood is abundant in the laminated dark and dark lithotypes within the M1B seam of the Latrobe Valley. Charred wood has a slightly heavier carbon isotopic composition when burnt at low temperatures (150–300 °C) (Jones and Chaloner, 1991; Czimczik et al., 2002; Ferrio et al., 2006). This enrichment is due to the different susceptibility to volatization of the two major components of wood (cellulose and lignin) (Ferrio et al., 2006). The increased abundance of charred wood in the laminated dark and dark lithotypes could therefore be contributing to the relatively heavy carbon isotope values of this lithotype. A forest bog, with a complete forest understorey, would contain a number of peat-forming plants that are deprived of sunlight and that grow in nutrient poor soils. Low light and low nutrient values result in lighter carbon isotope values in C3 plants (Farquhar et al., 1982), possibly contributing to the relatively light carbon isotope values in the medium dark lithotype. Isoprenoid lipids, such as those found in resins, characteristically have light carbon isotope values (Murray et al., 1998). Bechtel et al. (2008) also found that the carbon isotope composition of lignite is affected by the content of lipid-rich (resins, leaves, bark) organic matter. The abundance of resin in the medium dark lithotype (Holdgate et al., 2014) could therefore contribute to the light carbon isotopic values. Increased water stress can result in heavier carbon isotopic compositions as plants that experience water stress close their stomata to conserve water and increase water-use efficiency (Arens et al., 2000). The closed stomata inhibit water loss but also retard C02 gain, thereby producing heavier carbon isotope values (Brodribb, 1996). Holdgate et al. (2014) suggested that this process may have produced the heavier carbon isotopic values found in the lighter lithotypes. Angiosperms tend to have lighter carbon isotopic compositions than gymnosperms (on average 2.5‰ difference, Murray et al., 1998; Bechtel et al., 2000). In addition, gymnosperm plant material is more resistant to decay than angiosperm material (Poorter et al., 2009; Bond and Midgley, 2012). A possible explanation for the heavy carbon isotopic values in lighter lithotypes is therefore that lighter lithotypes have been subject to

249

greater levels of aerobic degradation, with preferential preservation of gymnosperm plant material. There has been a significant amount of research documenting the effects of aerobic degradation of organic matter on carbon isotopic compositions. In modern organic soils and peats, carbon isotopes of organic material generally become heavier with depth in the first several tens of centimetres (e.g. Natelhoffer and Fry, 1988; Santruckova et al., 2000; Krüger et al., 2015). Part of this trend appears to result from the addition of fossil fuel-derived carbon over the last two hundred years. However, this appears to only account for part of this heavying trend and several other mechanisms have been proposed as explanations, most relating to microbial breakdown of organic matter within aerobic environments (for a summary, see Boström et al., 2007). Santruckova et al. (2000) have suggested that microbial degradation processes become more important with higher temperatures. Given the interpreted subtropical nature of the Miocene Latrobe Valley palaeoclimate (Kershaw, 1997), it appears likely that microbial processes would be significant within the peatlands. An alternative explanation is that of variable mobility of dissolved organic carbon (Boström et al., 2007). Regardless of the precise mechanism involved, it is clear that lithotype and quantitative colourimetry have a strong correlation with carbon isotopes. Based on the new depositional environments assigned to each lithotype, new mechanisms for the resulting carbon isotopes of each respective lithotype can be estimated. Aerobic subsurface environments appear to produce heavy carbon isotopic compositions within peats. This being the case, we suggest that aerobic degradation of the peat is very likely the major cause of the heavying trend in carbon isotopes from the medium dark to pale lithotypes. We suggest that the lighter lithotypes have heavier carbon isotopic compositions because they were subjected to the most intense aerobic degradation. As suggested previously (Holdgate et al., 2014), the heavy values for the laminated dark lithotype are likely due to their original subaqueous depositional environment.

6.9. Controls on vegetation successions Extensive marshes develop during periods of relatively stable, high sea levels because this results in stable water tables (Anderson and Mackay, 1990; Holdgate et al., 2007). The well-developed lithotype cycles in the brown coal seams of Latrobe Valley were probably deposited during periods of relatively high sea level when stable water tables allowed the development of extensive marshes. All evidence from the Latrobe Valley coal cycles including geochemistry, micropaleontology and sedimentology suggest the dominance of drying-upwards cycles. Terrestrialization cycles are widely recognized in modern and ancient peatlands, with relative depth to water table as the principle driving-force. During periods of relative stability, a swamp undergoes the normal succession from an open-water setting, through a herbaceousdominated flora, to an ombrogenous raised-forested bog environment, reflecting the floral response to decreasing relative water table within the palaeomire/palaeoswamp (i.e. terrestrialization) (Moore and Bellamy, 1974; Frank and Bend, 2004). These cycles can be stacked (six cycles are recorded in the Palaeocene Souris Lignite) and reflect the peat system responding to changes in the ecological conditions in the mire (i.e. hydroseral succession, Frank and Bend, 2004). Potter et al. (2008) also proposed repeated terrestrialization (and drying) of ancient wetlands in the Lower Cretaceous Medicine River Coals of Alberta, where the brightening-upwards cycles are interpreted to represent “drying-upwards” cycles in terms of groundwater levels. Oscillations in the water table, due to eustatic sea level change or subsidence is also suggested to have triggered the process of vegetation succession in the brown coals of the German Lower Rhine Embayment and in the Niederlausitz Lignite District, Germany (Lücke et al., 1999; Seifert et al., 1993). The repeated cycles of terrestrialization recorded in the Latrobe Valley peatlands are therefore likely to reflect a series of drying-

250

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

upwards cycles as the marsh developed into an extensive ombrogonous raised forested bog. However, this normal progression of terrestrialization is punctuated by episodes of transgression (caused by eustatic sea level rise and subsidence), resulting in the termination of cycles (Frank and Bend, 2004). A relative rise in sea level, and consequently in the regional water table, would result in cycle termination by drowning. This mechanism appears to be responsible for the termination of peat accumulation directly below interseam sediments in the Latrobe Valley. The lowering of the perched water table in an ombrogenous raised bog, due to minor fluctuations in sea level would also result in the termination of cycles observed in coal seams. Terrestrialization cycles are essentially non-marine analogues of the well-established shallowing-upwards cycles found in marine successions (Holdgate et al., 2014).

7. Conclusions The shallowing-upwards (terrestrialization) cycles within the Latrobe Valley brown coals record a detailed floral and charcoal record from the Late Oligocene to Middle Miocene. Each of the lithotypes examined in detail from the #64 m lightening-upward cycle have a distinct floral component that is autochthonous and facies-controlled. The #64 m lightening upward cycle also records the development of a vegetation succession within the peatlands, resulting from variable depositional environments. The charcoal-rich laminated dark lithotype is interpreted to represent an emergent marsh with bulrushes that transitioned landward into a meadow marsh with rushes, heaths and coral ferns. The dark lithotype is interpreted as a transition from meadow marsh to forested bog, with members of the family Podocarpaceae representing the dominant trees. The medium lithotypes are interpreted to represent an ombrogenous forest bog with angiosperm taxa dominating the canopy. This depositional interpretation is generally consistent with, and refines the work of Holdgate et al. (2014), but for the first time, recognizes the significance of podocarps in the dark lithotype and angiosperms in the lighter lithotypes. The similarity between floral associations and successions in New Zealand and a lightening-upwards cycle from the Latrobe Valley coals suggests that New Zealand's vegetation communities represent a modern floral analogue for the vegetation successions in the Latrobe Valley, Oligo-Miocene coals. Moreover, like modern New Zealand, the Latrobe Valley peatlands were subject to repeated fires, as a single lightning strike in the Typhaceae-covered emergent marsh environment would probably ensure the entire marsh environment was burnt. All evidence from the Latrobe Valley coal cycles including geochemistry, micropaleontology, and sedimentology suggests the dominance of repeated drying-upwards or terrestrialization cycles. These well-developed lithotype cycles in the brown coal seams of Latrobe Valley were probably deposited during periods of relatively high sea level when stable, high water tables allowed the development of extensive marshes and vegetation successions.

Acknowledgements AGL Loy Yang and GHD are gratefully acknowledged for enabling the sampling of the M1B and M2A seams from Loy Yang Open Cut Mine. We are also grateful to Dr. Michael Shawn-Fletcher (University of Melbourne) for his assistance in developing methods used for palynological and charcoal processing, to Dr. Ian Sluiter, for his comments regarding the identification of Early Miocene palynomorphs and to Professor David Cantrill (Royal Botanic Gardens) for his insights into the identification of charcoalified floral fragments. We are also grateful to Cortland Eble, Mike Pole and one anonymous reviewer for their constructive and helpful comments on the manuscript.

References Anderson, J.A.R., 1964. The structure and development of the peat-swamps of Sarawak and Brunei. J. Trop. Geogr. 18, 7–16. Anderson, K.B., Mackay, G., 1990. A review and reinterpretation of evidence concerning the origin of Victorian Brown Coal. Int. J. Coal Geol. 16, 327–347. Anderson, J.A.R., Muller, J., 1975. Palynological study of a Holocene peat and a Miocene coal deposit from N.W. Borneo. Rev. Palaeobot. Palynol. 19, 291–351. Arens, N.C., Hope Jahren, A., Amundson, R., 2000. Can C3 plants faithfully record the carbon isotopic composition of atmospheric carbon dioxide? Paleobiology 26, 137–164. Attwood, D.H., Bolger, P.F., George, A.M., Murdoch, C.M., 1984. Measurement of the Colour of Dry Soft Brown Coal. State Electricity Commission of Victoria, Research and Development Department (SO/84/73). Baker, E.W., Louda, J.W., 1986. Porphyrins in the geological record. Mol. Gen. Genet. MGG 24, 125–225. Bechtel, A., Gratzer, R., Püttmann, W., Oszczepalski, S., 2000. Geochemical and isotopic composition of organic matter in the Kupferschiefer of the Polish Zechstein basin: relation to maturity and base metal mineralization. Int. J. Earth Sci. 89, 72–89. Bechtel, A., Gratzer, R., Sachsenhofer, R.F., Gusterhuber, J., Lücke, A., Püttmann, W., 2008. Biomarker and carbon isotope variation in coal and fossil wood of Central Europe through the Cenozoic. Palaeogeogr. Palaeoclimatol. Palaeoecol. 262, 166–175. Berggren, W.-A., Hilgen, F., Langereis, C., Kent, D.V., Obradovich, J., Raffi, I., Raymo, M.E., Shackleton, N., 1995. Late Neogene chronology: new perspectives in high-resolution stratigraphy. Geol. Soc. Am. Bull. 107, 1272–1287. Birks, H.J.B., 1981. The use of pollen analysis in the reconstruction of past climates: a review. In: Wigley, T.M., Ingram, M.J., Farmer, G. (Eds.), Climate and History: Studies in Past Climates and Their Impact on Man. Cambridge University Press, Cambridge, pp. 111–138. Birks, H.H., Whiteside, M., Stark, D.M., Bright, R., 1976. Recent paleolimnology of three lakes in northwestern Minnesota. Quat. Res. 6, 249–272. Blackburn, D.T., 1981. Floristic control on lithotype banding within the Yallourn coal seam, Victoria: evidence from megafossil assemblages, State Electricity Commission of Victoria unpublished report, Melbourne. Blackburn, D.T., 1985. Palaeobotany of the Yallourn and Morwell coal seams, State Electricity Commission of Victoria unpublished report, Melbourne. Blackburn, D.T., Sluiter, I.R.K., 1994. The Oligo-Miocene coal floras of southeastern Australia. In: Hill, R.S. (Ed.), History of Australian Vegetation: Cretaceous to Recent. Cambridge University Press, Cambridge, pp. 328–367. Bond, W.J., Midgley, G.F., 2012. Carbon dioxide and the uneasy interactions of trees and savannah grasses. Philos. Trans. R. Soc. B 367, 601–612. Bond, W.J., Dickinson, K.J., Mark, A.F., 2004. What limits the spread of fire-dependent vegetation? Evidence from geographic variation of serotiny in a New Zealand shrub. Glob. Ecol. Biogeogr. 13, 115–127. Boström, B., Comstedt, D., Ekblad, A., 2007. Isotope fractionation and 13C enrichment in soil profiles during the decomposition of soil organic matter. Oecologia 153, 89–98. Briggs, J., Large, D.J., Snape, C., Drage, T., Whittles, D., Cooper, M., Macquaker, J.H.S., Spiro, B.F., 2007. Influence of climate and hydrology in and early Miocene peatlands. Earth Planet. Sci. Lett. 253, 445–454. Brodribb, T., 1996. Dynamics of changing intercellular CO2 concentration (ci) during drought and determination of minimum functional ci. Plant Physiol. 111, 179–185. Brodribb, T., Hill, R.S., 2004. The rise and fall of the Podocarpaceae in Australia - a physiological explanation. In: Hemsley, A.R., Poole, I. (Eds.), The Evolution of Plant Physiology. Elsevier, Amsterdam, pp. 381–399. Bunting, M.J., Twiddle, C.L., Middleton, R., 2008. Using models of pollen dispersal and deposition in hilly landscapes: some possible approaches. Palaeogeogr. Palaeoclimatol. Palaeoecol. 259, 77–91. Cameron, C.C., Esterle, J.S., Palmer, C.A., 1989. The geology, botany and chemistry of selected peat-forming environments from temperate and tropical latitudes. Int. J. Coal Geol. 12, 105–156. Campbell, D.I., Williamson, J., 1997. Evaporation from a raised peat bog. J. Hydrol. 193, 142–160. Carpenter, R.J., Jordan, G.J., Lee, D.E., Hill, R.S., 2010. Leaf fossils of Banksia (Proteaceae) from New Zealand: an Australian abroad. Am. J. Bot. 97, 288–297. Carpenter, R.J., Bannister, J.M., Lee, D.E., Jordan, G.J., 2014. Nothofagus subgenus Brassospora (Nothofagaceae) leaf fossils from New Zealand: a link to Australia and New Guinea? Bot. J. Linn. Soc. 174, 503–515. Clark, R.L., 1982. Point count estimation of charcoal in pollen preparations and thin sections of sediments. Pollen Spores 24, 523–535. Clarkson, B.R., Schipper, L.A., Lehmann, A., 2004. Vegetation and peat characteristics in the development of lowland restiad peat bogs, North Island, New Zealand. Wetlands 24, 133–151. Cloern, J.E., Canuel, E.A., Harris, D., 2002. Stable carbon and nitrogen isotope composition of aquatic and terrestrial plants of the San Francisco Bay estuarine system. Limnol. Oceanogr. 47, 713–729. Collinson, M.E., 2002. The ecology of Cainozoic ferns. Rev. Palaeobot. Palynol. 119 (1), 51–68. Collinson, M.E., Steart, D.C., Scott, A.C., Glasspool, I.J., Hooker, J.J., 2007. Episodic fire, runoff and deposition at the Palaeocene-Eocene boundary. J. Geol. Soc. 164, 87–97. Conner, W.H., Sasser, C.E., Barker, N., 1986. Floristics of the Barataria Basin wetlands, Louisiana. Castanea 51, 111–128. Crawford, R.M., 2014. Tundra-Taiga Biology. Oxford University Press, pp. 99–100. Czimczik, C.I., Preston, C.M., Schmidt, M.W., Werner, R.A., Schulze, E.-D., 2002. Effects of charring on mass, organic carbon, and stable carbon isotope composition of wood. Org. Geochem. 33, 1207–1223. Deng, Y., Horrocks, M., Ogden, J., Anderson, S., 2006a. Modern pollen–vegetation relationships along transects on the Whangapoua Estuary, Great Barrier Island, northern New Zealand. J. Biogeogr. 33, 592–608.

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252 Deng, Y., Ogden, J., Horrocks, M., Anderson, S., 2006b. Application of palynology to describe vegetation succession in estuarine wetlands on Great Barrier Island, northern New Zealand. J. Veg. Sci. 17, 765–782. Diessel, C.F.K., 1992. Coal-bearing Depositional Systems. Springer-Verlag, Berlin, Heidelberg, pp. 1–721. Esterle, J.S., Ferm, J.C., 1994. Spatial variability in modern tropical peat deposits from Sarawak, Malaysia and Sumatra, Indonesia; analogues for coal. Int. J. Coal Geol. 26, 1–41. Farquhar, G.D., O'leary, M., Berry, J., 1982. On the relationship between carbon isotope discrimination and the intercellular carbon dioxide concentration in leaves. Funct. Plant Biol. 9, 121–137. Farquhar, G.D., Ehleringer, J.R., Hubick, K.T., 1989. Carbon isotope discrimination and photosynthesis. Annu. Rev. Plant Physiol. Plant Mol. Biol. 40, 503–537. Ferguson, D.K., Lee, D.E., Bannister, J.M., Zetter, R., Jordan, G.J., Vavra, N., Mildenhall, D.C., 2010. The taphonomy of a remarkable leaf bed assemblage from the Late Oligocene–Early Miocene Gore Lignite Measures, southern New Zealand. Int. J. Coal Geol. 83, 173–181. Ferrio, J., Alonso, N., López, J., Araus, J., Voltas, J., 2006. Carbon isotope composition of fossil charcoal reveals aridity changes in the NW Mediterranean Basin. Glob. Chang. Biol. 12, 1253–1266. Finotello, F., Johns, R.B., 1986. Some inter-relationships of kerogen and humic acid fractions in the Victorian brown coal pale lithotype. Org. Geochem. 9 (5), 265–273. Frank, M., Bend, S., 2004. Peat-forming history of the ancestral Souris mire (Palaeocene), Ravenscrag Formation, southern Saskatchewan, Canada. Can. J. Earth Sci. 41, 307–322. Gallagher, S., Smith, A., Jonasson, K., Wallace, M., Holdgate, G., Daniels, J., Taylor, D., 2001. The Miocene palaeoenvironmental and palaeoceanographic evolution of the Gippsland Basin, Southeast Australia: a record of Southern Ocean change. Palaeogeogr. Palaeoclimatol. Palaeoecol. 172, 53–80. George, A.M., Mackay, G.H., 1991. Petrology (chapter 2). In: Durie, R.A. (Ed.), The Science of Victorian Brown Coal: Structure, Properties and Consequences for Utilisation. Butterworth Heinemann, pp. 10–45. Gradstein, F.M., Ogg, J.G., Schmitz, M., Ogg, G., 2012. The Geologic Time Scale 2012, 2-Volume Set. Elsevier. Hagemann, H., Hollerbach, A., 1980. Relationship between the macropetrographic and organic geochemical composition of lignites. Phys. Chem. Earth 12, 631–638. Hagemann, H.W., Wolf, M., 1987. New interpretations of the facies of the Rhenish brown coal of West Germany. Int. J. Coal Geol. 7, 335–348. Holdgate, G.R., 2003. Coal, world-class energy reserves without limits. In: Birch, W.D. (Ed.), Geology of Victoria. Geological Society of Australia (Victoria Division), Special Publication 23, pp. 489–518. Holdgate, G.R., Kershaw, A.P., Sluiter, I.R.K., 1995. Sequence stratigraphic analysis and the origins of Tertiary brown coal lithotypes, Latrobe Valley, Gippsland Basin, Australia. Int. J. Coal Geol. 28, 249–275. Holdgate, G.R., Cartwright, I., Blackburn, D.T., Wallace, M.W., Gallagher, S.J., Wagstaff, B.E., Chung, L., 2007. The Middle Miocene Yallourn coal seam — the last coal in Australia. Int. J. Coal Geol. 70, 95–115. Holdgate, G.R., Wallace, M.W., Sluiter, I.R., Marcuccio, D., Fromhold, T.A., Wagstaff, B.E., 2014. Was the Oligocene–Miocene a time of fire and rain? Insights from brown coals of the southeastern Australia Gippsland Basin. Palaeogeogr. Palaeoclimatol. Palaeoecol. 411, 65–78. Holdgate, G., Wallace, M., O'Connor, M., Korasidis, V., Lieven, U., 2016. The origin of lithotype cycles in Oligo-Miocene brown coals from Australia and Germany. Int. J. Coal. Geol. (in press). Holloway, J.T., 1954. Forest and climates in the South Island of New Zealand. J. R. Soc. N. Z. 82, 329–410. Hunt, J., 2007. Wetlands of New Zealand: A Bitter-sweet Story. Random House, Auckland. HunterLab, 2007. HunterLab's guide to hitch standardization. Applications Note. 10(9). Jackson, S.T., Futyma, R.P., Wilcox, D.A., 1988. A paleoecological test of a classical hydrosere in the Lake Michigan dunes. Ecology 69, 928–936. Johnson, P.N., 2005. Fire in Wetlands and Scrub Vegetation: Studies in Southland, Otago, and Westland. Department of Conservation. Research and Development Series 215Department of Conservation, Wellington. Jones, T.P., Chaloner, W.G., 1991. Fossil charcoal, its recognition and palaeoatmospheric significance. Palaeogeogr. Palaeoclimatol. Palaeoecol. 97, 39–50. Jones, D.L., Clemesha, S.C., 1976. Australian Ferns and Fern Allies, With Notes on Their Cultivation. A.H. and A.W. Reed, Sydney. Kershaw, A., 1997. A bioclimatic analysis of early to middle Miocene brown coal floras, Latrobe Valley, south-eastern Australia. Aust. J. Bot. 45, 373–387. Kershaw, A.P., Sluiter, I.R.K., 1982. The application of pollen analysis to the elucidation of Latrobe Valley brown coal depositional environments and stratigraphy. Aust. Coal Geol. 4, 169–186. Kershaw, A.P., Bolger, P.F., Sluiter, I.R.K., Baird, J.G., Whitelaw, M., 1991. The nature and evolution of lithotypes in the Tertiary brown coals of the Latrobe Valley, southeastern Australia. Int. J. Coal Geol. 18, 233–249. Krüger, J., Leifeld, J., Glatzel, S., Szidat, S., Alewell, C., 2015. Biogeochemical indicators of peatland degradation–a case study of a temperate bog in northern Germany. Biogeosciences 12, 2861–2871. Leathwick, J.R., 1995. Climatic relationships of some New Zealand forest tree species. J. Veg. Sci. 6, 237–248. Leathwick, J., Austin, M., 2001. Competitive interactions between tree species in New Zealand's old-growth indigenous forests. Ecology 82, 2560–2573. Li, X., Rapson, G.L., Flenley, J.R., 2008. Holocene vegetational and climatic history, Sponge Swamp, Haast, south-western New Zealand. Quat. Int. 184, 129–138. Lücke, A., Helle, G., Schleser, G.H., Figueiral, I., Mosbrugger, V., Jones, T.P., Rowe, N.P., 1999. Environmental history of the German Lower Rhine Embayment during the Middle

251

Miocene as reflected by carbon isotopes in brown coal. Palaeogeogr. Palaeoclimatol. Palaeoecol. 154, 339–352. Luly, J., Sluiter, I.R.K., Kershaw, A.P., 1980. Pollen Studies of Tertiary Brown Coals: Preliminary Analysis of Lithotypes Within the Latrobe Valley, Victoria. Monash Publications in Geography 23. Mackay, G., Attwood, D.H., Gaulton, R.J., George, A.M., 1985. The cyclic occurrence of brown coal lithotypes. State Electricity Commission of Victoria Report SO/85/93, Melbourne. Macphail, M., 1999. Palynostratigraphy of the Murray Basin, inland southeastern Australia. Palynology 23, 197–240. Macphail, M., Alley, N., Truswell, E., Sluiter, I., 1994. Early Tertiary vegetation: evidence from spores and pollen. In: Hill, R.S. (Ed.), History of Australian Vegetation: Cretaceous to Recent. Cambridge University Press, Cambridge, pp. 189–261. Macphail, M., Field, L., Pillans, B., Davies, M., Hope, G., 2015. Lake George revisited: new evidence for the origin and evolution of a large closed lake, Southern Tablelands, NSW, Australia. Aust. J. Earth Sci. 62, 853–871. Mark, A., Smith, P., 1975. A lowland vegetation sequence in South Westland: Pakihi Bog to Mixed Beech-Podocarp Forest Part 1. The Principal Strata. NZ Ecol. Soc. 22, 76–92. Martin, J., Haider, K., Kirk, T., Higuchi, T., Chang, H., 1980. Microbial Degradation and Stabilization of 14C-labeled Lignins, Phenols, and Phenolic Polymers in Relation to Soil Humus Formation. CRC Press, Inc., pp. 77–100. McGlone, M.S., 1988. History of the New Zealand vegetation. In: Huntley, B., Webb, T. (Eds.), Vegetation History. Handbook of Vegetation Science 7. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 558–599. McGlone, M.S., 2009. Postglacial history of New Zealand wetlands and implications for their conservation. N. Z. J. Ecol. 33, 1–23. McGlone, M.S., Bathgate, J.L., 1983. Vegetation and climate history of the Longwood Range, South Island, New Zealand, 12 000 BP to the present. N. Z. J. Bot. 21, 292–315. McGlone, M., Neall, V., 1994. The late Pleistocene and Holocene vegetation history of Taranaki, North Island, New Zealand. N. Z. J. Bot. 32, 251–269. McK. Pegman, A.P., Ogden, J., 2005. Productivity-decomposition dynamics of Typha orientalis at Kaitoke Swamp, Great Barrier Island, New Zealand. N. Z. J. Bot. 43, 779–789. Mew, G., 1983. Application of the term “pakihi” in New Zealand—a review. J. R. Soc. N. Z. 13, 175–198. Mildenhall, D.C., 1980. New Zealand Late Cretaceous and Cenozoic plant biogeography: a contribution. Palaeogeogr. Palaeoclimatol. Palaeoecol. 31, 197–233. Mooney, S., Tinner, W., 2011. The analysis of charcoal in peat and organic sediments. Mires and Peat. 7, 1–18. Moore, P.D., Bellamy, D.J., 1974. Peatlands. Elek Science, London. Moore, P.D., Webb, J.A., 1978. Illustrated Guide to Pollen Analysis. Hodder and Stoughton. Moore, P.D., Webb, J.A., Collinson, M.E., 1991. Pollen Analysis. Blackwell Scientific Publications. Morley, R.J., 1981. Development and vegetation dynamics of a lowland ombrogenous peat swamp in Kalimantan Tengah, Indonesia. J. Biogeogr. 8, 383–404. Muller, S.D., Richard, P.J., Larouchel, A.C., 2003. Holocene development of a peatland (southern Québec): a spatio-temporal reconstruction based on pachymetry, sedimentology, microfossils and macrofossils. The Holocene 13, 649–664. Murray, A.P., Edwards, D., Hope, J.M., Boreham, C.J., Booth, W.E., Alexander, R.A., Summons, R.E., 1998. Carbon isotope biogeochemistry of plant resins and derived hydrocarbons. Org. Geochem. 29, 1199–1214. Natelhoffer, K., Fry, B., 1988. Controls on natural nitrogen-15 and carbon-13 abundances in forest soil organic matter. Soil Sci. Soc. Am. J. 52, 1633–1640. Newnham, R., 1992. A 30,000 year pollen, vegetation and climate record from Otakairangi (Hikurangi), Northland, New Zealand. J. Biogeogr. 19, 541–554. Nichols, G.J., Cripps, J.A., Collinson, M.E., Scott, A.C., 2000. Experiments in waterlogging and sedimentology of charcoal: results and implications. Palaeogeogr. Palaeoclimatol. Palaeoecol. 164, 43–56. Norton, D.A., 1989. Floristics and structure of mire-forest ecotones, west coast South Island, New Zealand. J. R. Soc. N. Z. 19, 31–42. Ogden, J., Stewart, G.H., 1995. Community dynamics of the New Zealand conifers. In: Enright, N.J., Hill, R.S. (Eds.), Ecology of the Southern Conifers. Melbourne University, Australia, pp. 81–119. Ogden, J., Stewart, G., Allen, R., 1996. Ecology of New Zealand Nothofagus forests. In: Veblen, T.T., Hill, R.S., Read, J. (Eds.), The Ecology and Biogeography of Nothofagus Forests. Yale University Press, pp. 25–82. Partridge, A.D., 2006. Late Cretaceous–Cenozoic palynology zonations Gippsland Basin. In Montiel, E., (co-ord) Australian Mesozoic and Cenozoic Palynology Zonations, Geoscience Australia, Record 2006/23. Pate, J.S., Meney, K.A., Dixon, K.W., Bell, T.L., Hickman, E.J., 1999. Australian rushes. In: Meney, K.A., Pate, J.S. (Eds.), Australian RushesAustralian Biological Resources Study. University of Western Australia Press. Patterson, W.A., Edwards, K.J., Maguire, D.J., 1987. Microscopic charcoal as a fossil indicator of fire. Quat. Sci. Rev. 6, 3–23. Perry, G.L., Wilmshurst, J.M., McGlone, M.S., 2014. Ecology and long-term history of fire in New Zealand. N. Z. J. Ecol. 38, 157–176. Poorter, H., Niinemets, Ü., Poorter, L., Wright, I.J., Villar, R., 2009. Causes and consequences of variation in leaf mass per area (LMA): a meta-analysis. New Phytol. 182, 565–588. Potter, J., PetroGraphics, J., McIlreath, I., Natras, T., 2008. Lithotypes, macerals and coal facies studies of Lower Cretaceous medicine river coals in South Central Alberta: applications in CBM exploration, depositional environments and tectonic history studies. Back to Exploration, pp. 777–780. Riegel, W., 2000. Two new doctoral theses on Eocene palynology in Germany. A.A.S.P. Newsletter, pp. 8–9. Robbins, R.G., 1962. The podocarp-broadleaf forests of New Zealand. Trans. Roy. Soc. NZ. Bot. 1, 34–75.

252

V.A. Korasidis et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 461 (2016) 237–252

Rochefort, L.I., Strack, M.A., Poulin, M.O., Price, J.S., Graf, M.A., Desrochers, A.N., Lavoie, C.L., 2012. Northern peatlands. In: Batzer, D.P., Baldwin, A.H. (Eds.), Wetland Habitats of North America: Ecology and Conservation Concerns. University of California Press, Berkeley, pp. 119–134. Rogers, G.M., Walker, S., Basher, L.M., Lee, W.G., 2007. Frequency and impact of Holocene fire in eastern South Island, New Zealand. N. Z. J. Ecol. 31, 129–142. Rydin, H., Jeglum, J., 2006. The Biology of Peatlands. Oxford University Press (1–343 pp.). Sander, P.M., Gee, C.T., 1990. Fossil charcoal: techniques and applications. Rev. Palaeobot. Palynol. 63, 269–279. Santruckova, H., Bird, M., Lloyd, J., 2000. Microbial processes and carbon-isotope fractionation in tropical and temperate grassland soils. Funct. Ecol. 14, 108–114. Scott, A.C., 2010. Charcoal recognition, taphonomy and uses in palaeoenvironmental analysis. Palaeogeogr. Palaeoclimatol. Palaeoecol. 291, 11–39. Scott, A.C., Cripps, J.A., Collinson, M.E., Nichols, G.J., 2000. The taphonomy of charcoal following a recent heathland fire and some implications for the interpretation of fossil charcoal deposits. Palaeogeogr. Palaeoclimatol. Palaeoecol. 164, 1–31. Seifert, A., Brause, H., Rascher, J., 1993. Geology of the Niederlausitz Lignite District, Germany. Int. J. Coal Geol. 23, 263–289. Sluiter, I.R.K., 1984. Palynology of Oligo-Miocene coal seams, Latrobe Valley, Victoria, PhD Thesis, Monash University, Melbourne (unpublished). Sluiter, I., Kershaw, A., 1982. The nature of Late Tertiary vegetation in Australia. Alcheringa 6, 211–222. Smith, G.C., 1982. A review of the Tertiary–Cretaceous tectonic history of the Gippsland Basin and its control on coal measure sedimentation. Aust. Coal Geol. 4, 1–38. Stewart, G., Basher, L., Burrows, L., Runkle, J., Hall, G., Jackson, R., 1993. Beech-hardwood forest composition, landforms, and soil relationships, north Westland, New Zealand. Vegetatio 106, 111–125. Stover, L., Partridge, A., 1973. Tertiary and Late Cretaceous spores and pollen from the Gippsland Basin, southeastern Australia. Proc. Roy. Soc. Victoria 85, 237–286. Taylor, D.J., 1966. Esso Gippsland Shelf No. 1, the mid-Tertiary foraminiferal sequence, in Appendix 2, Esso Gippsland Shelf No. 1 Well, Victoria. Department of Natural Development Bureau of Mineral Resources Geology and Geophysics Petroleum Search Subsidy Acts, Publication 76, pp. 31–46. Teichmüller, M., 1958. Metamorphisme du charbon et prospection du petrole. Rev. Ind. Minerale. Numero special, pp. 1–15.

Teichmüller, M., 1989. The genesis of coal from the viewpoint of coal petrology. Int. J. Coal Geol. 12, 1–87. Traverse, A., 1988. Paleopalynology. Unwin Hyman, Boston, pp. 1–600. Twiddle, C.L., Jones, R.T., Caseldine, C.J., Sugita, S., 2012. Pollen productivity estimates for a pine woodland in eastern Scotland: the influence of sampling design and vegetation patterning. Rev. Palaeobot. Palynol. 174, 67–78. Vaughan, A., Nichols, G.J., 1995. Controls on the deposition of charcoal: implications forsedimentary accumulations of fusain. J. Sediment. Res. A65 (1), 129–135. Verheyen, T.V., Johns, R.B., Bryson, R.L., Maciel, G.E., Blackburn, D.T., 1984. A spectroscopic investigation of the banding or lithotypes occurring in Victorian brown coal seams. Fuel 63, 1629–1635. Waddington, J.C., 1969. A stratigraphic record of the pollen influx to a lake in the Big Woods of Minnesota. Geol. Soc. Spec. Pap. 123, 263–282. Wade, B.S., Pearson, P.N., Berggren, W.A., Pälike, H., 2011. Review and revision of Cenozoic tropical planktonic foraminiferal biostratigraphy and calibration to the geomagnetic polarity and astronomical time scale. Earth-Sci. Rev. 104, 111–142. Walther, J., 1894. Einleitung in die Geologie als historische Wissenschaft: Beobachtungen über die Bildung der Gesteine und ihrer organischen Einschlüsse. G. Fischer. Wardle, P., 1964. Facets of the distribution of forest vegetation in New Zealand. N. Z. J. Bot. 2, 352–366. Wardle, P., 1974. The kahikatea (Dacrycarpus dacrydioides) forest of south Westland. N. Z. Ecol. Soc. 21, 62–71. Wardle, P., 1977. Plant communities of Westland National Park (New Zealand) and neighboring lowland and coastal areas. N. Z. J. Bot. 15, 323–398. Wardle, P., 1980. Primary succession in Westland national park and its vicinity, New Zealand. N. Z. J. Bot. 18, 221–232. Williams, M.A.J., Dunkerley, D.L., De Decker, P., Kershaw, A.P., Stokes, T.J., 1993. Quaternary Environments. Edward Arnold, London, pp. 57–201. Wilmshurst, J.M., McGlone, M.S., Charman, D.J., 2002. Holocene vegetation and climate change in southern New Zealand: linkages between forest composition and quantitative surface moisture reconstructions from an ombrogenous bog. J. Quat. Sci. 17, 653–666. Winkler, E., 1986. Organic geochemical investigations of brown coal lithotypes. A contribution to facies analysis of seam banding in the Helmstedt deposit. Org. Geochem. 10, 617–624.