bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Messinian vegetation and climate of the intermontane Florina-Ptolemais-Servia Basin,
NW Greece inferred from palaeobotanical data: How well do plant fossils reflect past
environments?
Johannes M. Bouchal1*, Tuncay H. Güner2, Dimitrios Velitzelos3, Evangelos Velitzelos3,
Thomas Denk1
1
Swedish Museum of Natural History, Department of Palaeobiology, Box 50007, 10405
Stockholm, Sweden
2
Faculty of Forestry, Department of Forest Botany, Istanbul University Cerrahpaşa, Istanbul,
Turkey
3
National and Kapodistrian University of Athens, Faculty of Geology and Geoenvironment,
Section of Historical Geology and Palaeontology, Greece
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
The late Miocene is marked by pronounced environmental changes and the appearance of
strong temperature and precipitation seasonality. Although environmental heterogeneity is to
be expected during this time, it is challenging to reconstruct palaeoenvironments using plant
fossils. We investigated leaves and dispersed spores/pollen from 6.4–6 Ma strata in the
intermontane Florina-Ptolemais-Servia Basin (FPS) of NW Greece. To assess how well plant
fossils reflect the actual vegetation of the FPS, we assigned fossil-taxa to biomes providing a
measure for environmental heterogeneity. Additionally, the palynological assemblage was
compared to pollen spectra from modern lake sediments to assess biases in spore/pollen
representation in the pollen record. We found a close match of the Vegora assemblage with
modern Fagus–Abies forests of Turkey. Using taxonomic affinities of leaf fossils, we further
established close similarities of the Vegora assemblage with modern laurophyllous oak forests
of Afghanistan. Finally, using information from sedimentary environment and taphonomy, we
distinguished local and distantly growing vegetation types. We then subjected the plant
assemblage of Vegora to different methods of climate reconstruction and discussed their
potentials and limitations. Leaf and spore/pollen records allow accurate reconstructions of
palaeoenvironments in the FPS, whereas extra-regional vegetation from coastal lowlands is
likely not captured.
Keywords
Biome reconstruction, proxy biases, climate reconstruction, plant macrofossils, dispersed
pollen, light and scanning electron microscopy
1. Introduction
The late Miocene (11.6–5.3 Ma) marks the time in the Neogene (23–2.58 Ma) with the largest
shift from equable climate to strong latitudinal temperature gradients in both hemispheres
(Herbert et al., 2016). This is well illustrated by the global rise of C4 dominated ecosystems
(grasslands and savannahs in the tropics and subtropics; Cerling et al., 1997). In the
Mediterranean region, vegetation changes did not happen synchronously with modern steppe
and Mediterranean sclerophyllous woodlands replacing humid temperate forest vegetation at
different times and places during the middle and late Miocene (Suc, 1984; Kovar-Eder, 2003;
Fauquette et al., 2006; Kovar-Eder et al., 2014; Bouchal et al., 2018; Denk et al., 2018; Suc et
al., 2018). During the latest Miocene (5.9–5.3 Ma) the desiccation of the Mediterranean Sea
was caused by the isolation of the Mediterranean Sea from the Atlantic Ocean. Based on
palynological studies across the Mediterranean region, this event did not have a strong effect
on the existing vegetation. Open and dry environments existed in southern parts before,
during, and after this so-called Messinian Salinity Crisis (MSC). In contrast, forested
vegetation occurred in northern parts of Spain, Italy, and the Black Sea (Fauquette et al.,
2006). Likewise, a vegetation gradient occurred from N and C Italy and Greece to Turkey,
where humid temperate forests had disappeared by the early late Miocene (Denk et al., 2018).
The Florina-Ptolemais-Servia Basin (FPS) of NW Greece is one of the best-understood
intermontane basins of late Miocene age in the entire Mediterranean region. A great number
of studies investigated the tectonic evolution, depositional history, and temporal constraints of
basin fills (e.g. Steenbrink et al. 1999, 2000, 2006), plant fossils (e.g. Knobloch & Velitzelos,
1986a,b; Velitzelos & Gregor, 1985; Mai & Velitzelos, 1992; Kvaček et al., 2002; Velitzelos
et al., 2014), and vertebrate fossils (van de Weerd, 1979; Koufos 1982, 2006; Koufos et al.,
1991).
2
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
The Messinian flora of Vegora in the northern part of the FPS is dated at 6.4–6 Ma and
represents the vegetation in this region just before the onset of the Messinian salinity crisis
(the pre-evaporitic Messinian). This flora has been investigated since 1969 (Kvaček et al.,
2002) and represents one of the richest late Miocene leaf floras in the eastern Mediterranean
along with two other, slightly older, Messinian plant assemblages from the FPS,
Likoudi/Drimos and Prosilio/Lava (4-6 in Fig. 1; Velitzelos et al., 2014). The focus of
previous palaeobotanical studies in the FPS has been on macrofossils. In contrast, no
comprehensive study of dispersed pollen and spores has been carried out in the FPS. While
fruit and seed floras, to a great extent, and leaf floras, to a lesser extent, reflect local
vegetation in an area, dispersed pollen and spores provide additional information about the
regional vegetation. Therefore, a main focus of the present study is on spores and pollen of
the Messinian plant assemblage of Vegora.
We (i) investigated dispersed spores and pollen using a combined light and scanning electron
microscopy approach (Daghlian, 1982; Zetter, 1989; Halbritter et al., 2018) that allows a
more accurate determination of pollen and hence higher taxonomic resolution. We (ii) then
compiled a complete list of plant taxa recorded for the site of Vegora including fruits and
seeds, foliage, and spores and pollen. Based on the ecological properties of their modern
analogue taxa, we assigned the fossil-taxa to functional types (vegetation units) and inferred
palaeoenvironments of the FPS during the Messinian. Using leaf physiognomic
characteristics, we (iii) conducted a Climate Leaf Analysis Multivariate Program (CLAMP)
analysis (Spicer, 2008; Yang et al., 2011) to infer several climate parameters for the late
Miocene of the FPS. We also (iv) used a modified “Co-existence approach” (Mosbrugger &
Utescher, 1997) based on climatic requirements of modern analogue plant taxa to infer two
climate parameters and (v) a Köppen signature analysis (Denk et al., 2013; Bouchal et al.,
2018) based on the Köppen-Geiger climate types in which modern analogue taxa of the fossiltaxa occur. Finally, we (vi) discuss how well the translation of fossil plant assemblages into
functional types (vegetation units, biomes) works for reconstructing past environments at
local and regional scales.
2. Material and Methods
2.1. Geological setting
The old open-pit lignite quarry of Vegora is located in western Macedonia, NW Greece, ca. 2
km E of the town of Amyntaio and is part of the Neogene Florina-Ptolemais-Servia
intermontane basin (FPS). The FPS is part of the Pelagonian basin that extends to the north
into North Macedonia (Fig. 1). The NNW–SSE trending FPS is ca. 120 km long and presently
at elevations between 400 and 700 m a.s.l. and is flanked by mountain ranges to the east and
the west. Main ranges include Baba Planina (2,601 m), Verno (2,128 m), and Askio (2,111 m)
to the east of the basin and Voras (2,528 m), Vermio (2,065 m), Olympus (2,917 m) to the
west (Fig. 1). These ranges are mainly comprised of Mesozoic limestones, Upper
Carboniferous granites and Paleozoic schists.
Continuous sedimentation since 8 Ma resulted in the accumulation of ca. 600 m of late
Miocene to early Pleistocene lake sediments with intercalated lignites and alluvial deposits.
The FPS basin formed in the late Miocene as a result of NE–SW extension in the Pelagonian
Zone, the westernmost zone of the Internal Hellenides (Brunn, 1956; Pavlides & Mountrakis,
1987). A subsequent Pleistocene episode of NW–SE extension caused the fragmentation of
the basin into several sub-basins (Steenbrink et al., 2006).
3
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
Basin fills overlay unconformably Paleozoic and Mesozoic rocks. Alpine and pre-Alpine
basement of the area consists of Pelagonian metamorphic rocks (gneisses, amphibolites, mica
schists, meta-granites and Permian to Triassic meta-sediments) and crystalline limestone of
Triassic–Jurassic age (carbonate cover). Subpelagonian ophiolites and deep-sea sediments of
Jurassic age, comprising the Vourinos ophiolitic complex, thrust over the Pelagonian
carbonate rocks and are covered by Cretaceous strata (Brunn, 1956; Mavridis et al., 1977;
Mountrakis, 1986; Roberts et al., 1988).
The Vegora section belongs to the ca. 300 m thick Komnina Formation, which
unconformably overlies pre-Neogene basement and is predominantly composed of alluvial
sands and conglomerates, lacustrine (diatomaceous) marls and palustrine clays, with some
intercalated (xylite-type) lignite seams (Steenbrink et al., 2006). The detailed description of
the sequence at the Vegora quarry follows Kvaček et al. (2002) and Steenbrink et al. (2006)
(Fig. 2). Since 2000, the lower part of the sequence was not accessible and the exposed
sequence started with a ca. 10 m thick lignite seam (see fig. 3 in Kvaček et al., 2002 versus
fig. 2, unit 1, in Steenbrink et al., 2006).
The full Vegora section begins with hard marls (>15 m) followed by 10–15 m of clay sands
and a white marl layer of 10 m. Then, a formation of clay sands follows with a total thickness
of 15–20 m. This formation starts with lignitic marls followed by marls and clay sand
intercalations. The sand is rich in mica.
Above this, a lignite seam occurs with a thickness of 12–35 m. Within the seam, three xylitic
layers with a total thickness of 10–12 m can be distinguished. The lower xylitic layer is about
3–4 m, the middle 0.5–1.5 m and the upper is 4–9 m in thickness (Kvaček et al., 2002).
The upper lignite layer was the first visible layer in 2002 in the section (unit 1 in Steenbrink et
al., 2006). Between the xylitic layers, sand layers of various thicknesses (0 to 15 m) occur. In
general, the thickness of these sand layers is smaller towards the N and NE ends of the mine
and becomes significantly larger towards the W and SW ends of the mine.
The top of the upper xylitic layer is covered by 3–4 m thick shales, followed by a thick layer
of light blue marls, 10–60 m thick (unit 2 in Steenbrink et al., 2006), and a layer of sandy
marls, 10–40 m thick (units 3 and 4 in Steenbrink et al., 2006). Unit 3 is made up of greybrown lignitic clay at its base and multi-coloured mottled clays, silts and fine sands higher up.
Unit 4 consists of cross-bedded conglomerates and coarse sands at the base overlain by
mottled clays, silts and fine sands with calcareous nodules. Finally, the top of the section is
made up of dark red, mottled sands, silts and clays (unit 5).
Unit 2 of Steenbrink et al. (2006) corresponds to the main fossiliferous layers for plant fossils
and diatoms (Gersonde & Velitzelos, 1978).
The uppermost layer of the Neogene sediments in the area is a formation of marly limestones,
of different thickness, which is not everywhere visible appearing only at the nearby villages
of Neapoli and Lakia. All Neogene sediments of the area are inclined by a 10º slope towards
NNW.
The rocks on the top of the Neogene section are Quaternary alluvial deposits, conglomerates,
sands, and gravels. This material, in general, has been supplied from erosion processes of the
nearby metamorphic mountains.
2.2. Age
In the upper part of unit 2, 4 cm-thick layers of tephra rich in biotite were found and used for
40
Ar/39Ar dating (Steenbrink et. al., 2006). The calculated age of 5.97 ±0.07 Ma corresponds
to Messinian. Also, using palaeomagnetic data, the base of unit 2, just above the lignite seam,
4
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
can be correlated to the astronomical polarity timescale, indicating that its position
corresponds to the end of chron C3An.2n. This would suggest an age of ca. 6.4 Ma for the
beginning of unit 3. Therefore, the period of deposition of the light blue marls (unit 2) from
Vegora can be narrowed down to ca. 400 ka.
2.3. Sample processing
The palynological sample was taken from a slab piece (S115992) from a leaf layer in unit 2 of
the Vegora mine. The sample was processed following standard protocols (20% HCl to
dissolve carbonate, 40% HF to dissolve silica, 20% HCl to dissolve fluorspar; chlorination,
acetolysis; see Halbritter et al., 2018) and the residue was transferred to glycerol.
2.4. Palynological investigation
Light microscopy (LM) micrographs were taken with an Olympus BX51 microscope
(Swedish Museum of Natural History [NRM], Stockholm, Sweden) equipped with an
Olympus DP71 camera. The same grains were examined using LM and scanning electron
microscopy (SEM; single grain method; Zetter, 1989; Halbritter et al, 2018). Specimens were
sputter-coated with gold for SEM investigation. SEM micrographs were taken with an ESEM
FEI Quanta FEG 650 SEM (Stockholm University). Residue and SEM stubs are stored at
NRM under specimen numbers S11599201–S11599220. Terminology of palynomorphs
follows Punt et al. (2007) and Halbritter et al. (2018). Size categories follow Halbritter et al.
(2018). Palynomorphs were determined to family, genus or infrageneric level. In cases when
no taxonomic affinity could be established, we used fossil form taxa which are not implying a
particular systematic affiliation. The systematic palaeobotany section starts with algae, fern
and fern allies, gymnosperms and is followed by angiosperms. Angiosperm classification and
author names of orders and families follow APG IV (2016).
2.5. Inferring palaeoclimate estimates
We employed three different (semi)quantitative methods to infer a range of climate
parameters for the Messinian of NW Greece.
CLAMP (climate leaf analysis multivariate program) is a physiognomy based, taxon free
method of climate inference and makes use of the relationship between leaf architecture and
climate. CLAMP uses calibration datasets of modern vegetation sites across the world to
place a fossil leaf assemblage in physiognomic space, which then can be translated into
numeric values for several climate parameters (Spicer, 2008; Yang et al., 2011).
The Coexistence Approach (CA; Utescher et al., 2014) is a method of inferring palaeoclimate
based on nearest living relatives (NLR) of fossil taxa. CA assumes that for a given climate
parameter, the tolerances of all or nearly all taxa in a fossil assemblage will overlap to some
degree; this overlap is called the climatic coexistence interval. Following best practices in
applying the CA Utescher et al. (2014) provided several guidelines to apply the CA in a
meaningful way. Among these guidelines, one is to exclude relict taxa (usually monotypic or
comprising very few extant species) from the analysis, because of their likely
unrepresentative modern distribution. Examples for such taxa are the East and SE Asian
Craigia and Glyptostrobus. These taxa had a much wider distribution during parts of the
Cenozoic including Arctic regions. For example, Budantsev & Golovneva (2009) described
Craigia from the Eocene Renardodden Formation of Spitsbergen for which they inferred a
mean annual temperature (MAT) of 8.4 °C and a coldest month mean temperature (CMMT)
of –1 °C. In contrast, the two modern species occur in climates with MAT 13.2–21 °C and
5
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
CMMT 6.3–14.2 °C (Fang et al., 2009). Hence, it is assumed that for relict plants ecological
niches may have changed considerably during the Cenozoic. For further assumptions of the
CA and their critique, see Grimm & Potts (2016) and Grimm et al. (2016). Climate
parameters for the NRL are given in the electronic supplementary material Table S1.
Köppen signatures (Denk et al., 2015; Bouchal et al., 2018) is another approach to infer largescale climatic patterns for the Cenozoic that is based on NLR of fossil taxa. Modern
distribution ranges are mapped on Köppen-Geiger climate maps (Kottek et al., 2006; Peel et
al., 2007; Rubel et al, 2017; Global_1986-2010_KG_5m.kmz; definition of the used KöppenGeiger categories are given in the electronic supplementary material S6) and the Köppen
climate types in which the modern taxa occur are taken as a proxy for the climate space in
which the fossil taxa occurred. It is explicitly stated that climate niche evolution will
negatively impact the reliability of the inferred palaeoclimate. To overcome this drawback,
subgenera, sections, and genera are used as NLR, whereas single species are usually not
considered for NLR. The representation of different climate types is first scored for each
species within a genus as present (1)/absent (0) (electronic supplementary material Table S2).
To summarize preferences for climate types of all modern analogues, an implicit weighting
scheme is used to discriminate between modern analogues that are highly climatically
constrained and those that occur in many climate zones. For each modern species, the sum of
its Köppen signature is always 1. For example, if a species is present in two Köppen–Geiger
climate types, Cfa and Cfb, both score 0.5. If a species is present in 10 Köppen–Geiger
climate types, each of these climate types scores 0.1. The Köppen signature of a genus or
section, the preferred NLR of a fossil taxon, is the sum of its species’ Köppen signatures for
each climate type divided by the total number of scored species for this genus. By this, the
percentage representation of each Köppen–Geiger climate type is determined for a
genus/section (Bouchal et al., 2018). For pollen taxa of herbaceous and a few woody
angiosperm groups that are resolved to family-level only, the distributions of extant members
of the family were combined into a general family distribution range and the corresponding
Köppen–Geiger climate types determined.
2.6. Characterisation of terrestrial biomes
For convenience, we use the biome classification of Woodward et al. (2004) that recognizes
five major tree biomes based on the physiognomy of the dominant species: Needleleaf
evergreen (NLE), needleleaf deciduous (NLD), broadleaf evergreen (BLE), broadleaf cold
deciduous and broadleaf drought deciduous (BLDcold, BLDdrought). These authors also
observed that broadleaf drought deciduous vegetation grades substantially into broadleaf
evergreen vegetation. Besides, shrublands are defined as lands with woody vegetation less
than 2 m tall. Savannahs are defined as lands with herbaceous or other understorey systems,
where woody savannahs have forest canopy cover between 30 to 60 %, and savannah has
forest canopy cover between 10 and 30 % (Woodward et al., 2004). This very broad definition
of savannah may be strongly oversimplified. Thus, for savannah-like vegetation, we make a
distinction between steppe and forest-steppe of temperate regions with a continuous layer of
C3 grasses and savannah and woody savannah of tropical regions with a continuous layer of
C4 grasses (Ratnam et al., 2011).
3. Results
3.1. Pollen and spores: diversity and environmental signal
We determined more than 50 palynomorph taxa from a leaf layer in unit 2 of the Vegora
6
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
section (Figs 3–5). A comprehensive taxonomic account including pollen morphological
descriptions is provided in the electronic supplementary material S3. The fossil-taxa comprise
two algae, five ferns, twelve herbaceous plants, one woody liana, and more than 30 woody
trees and shrubs (Table 1). Besides the taxonomic evaluation, 430 palynomorphs were
counted to assess the abundance of different taxonomic groups, life forms, and pollination
syndromes. Roughly, half of the pollen taxa were present in very small amounts (1–3 grains
in the counted sample or < 1%; electronic supplementary material Table S4). The presence
and abundance of Spirogyra zygospores/ aplanospores indicate a lake with shallow lake
margins (reed belt with Typha) and stagnant, oxygen-rich, open freshwaters. Spores of
Osmunda (> 4%) and Leavigatosporites haardti (> 1.6%) are of moderate abundance
suggesting that the producing pteridophytes grew close to the sedimentation area, the
Messinian Vegora Lake. Among conifers and wind-pollinated trees, strong pollen producers
such as Pinus (subgenus Strobus 11%, subgenus Pinus ~8%), Abies (8.4%), Cathaya (7%)
and the Betulaceae Alnus (9.3%) and the Fagaceae Fagus (7%) are most abundant. Another
group of wind-pollinated woody trees and shrubs was represented with abundances between
2% and 5%. Among these were both deciduous and evergreen oaks (Quercus) and conifers
such as Cedrus and Tsuga and undifferentiated papillate Cupressaceae.
Among the taxa that are represented by single or few pollen grains, a significant number
belonged to insect-pollinated plants. Insect-pollinated trees, shrubs and lianas include
Craigia, Platycarya, Castaneoideae and Parthenocissus; Hedera and Sassafras are further
insect-pollinated woody taxa, recorded in the leaf fossil record. Herbaceous taxa comprise
Apiaceae, Caryophyllaceae, Geranium, Succisa, and Cichorioideae.
The ratio arboreal pollen (AP) to non-arboreal pollen (NAP) is 89.5% to 10.5% indicating a
forest dominated (tree prevalent) local and regional vegetation according to the threshold
values of Favre et al. (2008). Forest types (biomes of Woodward et al., 2004) represented by
the pollen assemblage are needleleaf evergreen and deciduous forests (NLE, NLD), broadleaf
deciduous forests (BLD), broadleaf evergreen forests (BLE) and mixed forests (MIXED). In
addition, BLD and NLD either thrived on well-drained soils or in temporally or permanently
inundated areas.
A few taxa might also indicate the presence of closed or open shrublands and grasslands
(herbaceous taxa including sparse Poaceae with affinity to Poa/Lolium, Chenopodiaceae,
Apiaceae etc. and woody taxa including palms; see Table 1 for other woody taxa). These may
have been associated with BLE woodlands (Q. mediterranea, Q. sosnowskyi) or with
mesophytic evergreen forests of Q. drymeja (see below). Alternatively, they may have
originated from an independent vegetation type (for example montane grasslands).
Among needleleaf forest biomes, for some taxa, the attribution to a distinct forest type is not
straightforward. For example, conifers such as Cathaya may have been part of the montane
hinterland vegetation on well-drained soils but may also have been important elements of
peat-forming vegetation (Schneider, 1992; Dolezych & Schneider, 2006, 2007, 2012).
Based on pollen abundances (electronic supplementary material Table S4) local (close to the
lake), regional (occurring in the FPS) and extra-regional (potentially occurring outside the
FPS) vegetation can be inferred. Regional vegetation consisted of BLD forests subjected to
flooding (Alnus) and NLD swamp forests (papillate Cupressaceae). Close to the lake, a mixed
forest with Fagus, Abies and Cathaya thrived (using the modern Abant Gölü of northern
Turkey as a reference for pollen rain vegetation relationships; van Zeist & Bottema, 1991).
Deciduous oaks (mainly of sect. Cerris) also might have been part of local forest vegetation
(BLD). Pinus and Cedrus NLE forests and evergreen oak forests (BLE) grew at some
7
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
distance from the lake (regional vegetation; using threshold abundances of Cedrus, 7%, and
evergreen Quercus, 20%, as indicators of local source vegetation; Bell & Fletcher, 2016).
3.2. Leaf fossils, fruits, and seeds: diversity and environmental signal
Leaf and fruit remains from the Vegora mine have been collected and described since half a
century (Schneider & Gregor, 1973; Velitzelos & Gregor, 1985, 1990; Knobloch &
Velitzelos, 1987, Mai & Velitzelos, 1997). The most recent reviews are those of Kvaček et al.
(2002) and Velitzelos et al. (2014). Table 1 provides an updated taxon list. Fruits and seeds
recovered from the lignite seam of the Vegora section represent aquatic and reed vegetation.
From the lignite seam also trunks of tall trees (as Sequoioxylon) in upright position were
recovered. The plant assemblage of the blue marls represents needleleaf evergreen and
deciduous, and broadleaf evergreen and deciduous, as well as mixed forests. Among
needleleaf forests, swamp forests are typically represented by Taxodioideae, while
Sequoioideae, Keteleeria, Pinus and others may have been part of peat-forming vegetation
(forests types NLE, NLD). Based on the great abundance of Alnus leaves, a local alder swamp
may also be inferred (BLD). Fagus is among the most abundant taxa based on the number of
recovered leaf remains suggesting that it was part of the mesic forest vegetation close to the
lake. Also, deciduous foliage of Quercus sect. Cerris (Q. kubinyi, possibly Q. gigas) might
have grown in the vicinity of the lake, either forming mixed stands with Fagus or oakdominated forests. Kvaček et al. (2002) referred to this vegetation as Fagetum
gussonii/Quercetum mixtum.
Evergreen oaks are abundant in the Vegora leaf assemblage but fairly rare in the pollen record
(electronic supplementary material Table S4). This indicates that the leathery leaves of these
taxa were transported to the area of sedimentation by slow-flowing streams and that the
source vegetation was further away from the lake. Kvaček et al. (2002) referred to these
evergreen forests as sclerophyllous (Quercetum mediterraneum). Denk et al. (2017)
distinguished between sclerophyllous Mediterranean oak forest and laurophyllous Q.
floribunda forest from Afghanistan which is a better analogue for the widespread western
Eurasian fossil-taxon Q. drymeja. Hence we infer an ecological cline from mesic evergreen
oak forests to sclerophyllous forest and shrublands in the Messinian of Vegora (cf. Freitag,
1982).
Well-drained forests dominated by needleleaf taxa occurred in the montane vegetation belt
(Abies) and on rocky substrates (Cedrus, Pinus). Only a few taxa are potentially representing
open shrubland vegetation (Acer spp., Chamaerops).
3.3. Inferring past climate with CLAMP
Forty-one dicot leaf morphotypes were scored for the CLAMP analysis. Given the distinctly
temperate appearance of this flora, we used the calibration dataset
Physg3arcAZ_GRIDMet3arAZ. Physg3arcAZ includes 173 sites, among them the 144
Physg3brcAZ sites plus 29 sites corresponding to the alpine nest (Wolfe, 1993). The alpine
are the coldest sites known to have a different physiognomic behaviour (Wolfe, 1993); they
are characterised by a WMMT lower than 16 °C and a CMMT lower than 3 °C (Wolfe, 1993).
The reconstructed climate parameters are MAT 10–13.5 °C, WMMT 19.2–22.8 °C, CMMT
1–5 °C, GROWSEAS 6–8 months, GSP 700–1100 mm, MMGSP 110–160 mm, Three_WET
500–780 mm, Three_DRY 180–260 mm, and Three_WET to Three_DRY ratio < 4. In terms
of the Köppen-Geiger climate classification, this translates into a temperate Cfb climate (Tcold
> 0 and < 18 °C; without a dry season; warm summer Thot < 22 °C).
8
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
In addition, we used the calibration dataset PhysgAsia1_HiResGRIDMetAsia1 that adds 45
sites from China to the Physg3brcAZ dataset. Using this dataset, the reconstructed climate
parameters are generally cooler and drier than the ones obtained from Physg3arcAZ. MAT
8.7–11.5 °C, WMMT 19–22.6 °C, CMMT -2.7–2.3 °C, GROWSEAS 5.5–7 months, GSP
460–1100 mm, MMGSP 100–160 mm, Three_WET 400–750 mm, Three_DRY 80–220 mm,
and Three_WET to Three_DRY ratio < 5.5. In terms of the Köppen-Geiger climate
classification, this translates into a temperate Cfb to cold Dfb climate (Tcold > 0 and < 18 °C
versus Tcold < 0 °C; both without a dry season and warm summer Thot < 22 °C). Score sheets
and full documentation of the CLAMP analyses are provided in electronic supplementary
material S5.
3.4. Inferring past climate with CA
Using the coexistence approach, we estimated coldest month mean temperature (CMMT) and
mean annual temperature (MAT) coexistence intervals to see how CA behaves including and
excluding relictual and monotypic taxa. Following Utescher et al. (2014) relict taxa with very
limited modern distribution were excluded from the analysis. Excluded taxa are plotted to the
left of the diagram in Fig. 6. For the monotypic genus Chamaerops, the tribus Trachycarpeae
was used as NLR. For CMMT a lower boundary value of 1.2 °C is estimated based on the
cold tolerance of Trachycarpeae. Chamaerops has a slightly warmer CMMT of 4 °C. For
mean annual temperature (MAT) the lower boundary is defined by Zelkova (8.6 °C) and the
upper boundary by Acer sect. Acer (21.2 °C). When only the 10–90% percentiles were
considered, MAT low was defined by Zelkova as 9.9 °C and MAT high by Acer sect. Acer as
18.4 °C (Fig. 6).
Inclusion of a priori excluded relict species with a limited distribution would greatly change
the estimated climate values. CMMT low would be defined by Craigia (6.5 °C) and CMMT
high by Sequoia (7.5 °C). Likewise, MAT low would be defined by the monotypic conifer
Cathaya (13.4 °C) and MAT high again by Sequoia (15.3 °C; electronic supplementary
material Table S1).
Using only the 10–90% percentiles, MAT low would be defined by Craigia (14 °C) and MAT
high (14.1 °C) by Sequoia.
3.5. Inferring past climate with Köppen signatures
Based on 700 Köppen signatures of modern species (rarely sections and families) genus- to
family-specific Köppen signatures were used to generate Köppen signatures for the Vegora
assemblage of unit 2. Temperate C climates are by far the most common ones represented by
modern analogues of the Vegora plant assemblage. Cfa/b and Cwa/b climates represent 50%
of the occurrences of NLR taxa when pollen and spores are considered, and 54% when
macrofossils are considered (Fig. 7). Csa/b climates are represented by 11–13%. Snow
climates (CMMT < 0°C) are represented by 17% (Df, Dw) and 2–3% (Ds). Thus, C and D
climates make up > 80% of all NLR occurrences. In contrast, equatorial climates are
represented by 10% (spores and pollen) and 6.5% (macrofossils). Arid B climates are
represented by < 10% in the spores/pollen and macrofossil assemblages.
4. Discussion
4.1. How well do plant fossils reflect past environments of the FPS?
It has long been known that there is no exact relationship between fossil (and modern)
assemblages of dispersed spores and pollen and the actual vegetation (e.g. van Zeist et al.,
9
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
1975; Connor et al., 2005; Traverse, 2007; Bell & Fletcher, 2016; Marinova et al., 2018).
Marinova et al. (2018) pointed out several problems when inferring vegetation from pollen
diagrams. These included (i) pollen production biases which generally result in the overrepresentation of woody species and the under-representation of herbaceous species in the
pollen assemblage, (ii) transport of tree pollen into non-forested areas resulting in poor
delineation of ecotonal boundaries, (iii) upslope transport of pollen from lowland areas in
upland areas resulting in poor delineation of altitudinal vegetation gradients and tree line.
Furthermore, these authors found that samples from small basins (<1 km2) are more likely to
be reconstructed accurately because they sample an appropriate pollen source area to reflect
regional vegetation patterns in relatively heterogeneous landscapes. In contrast, large
uncertainties were observed when inferring the local vegetation in large basins, e.g. the Black
Sea. Here, large pollen source areas result in strongly mixed signals which do not well
discriminate the vegetation belts around a specific site.
We note that this caveat may in fact be beneficial when inferring the past vegetation in a
larger area. The FPS is a basin that extends ca. 120 km x 30 km and is flanked by high
mountains. Hence, a rich pollen assemblage with a strongly mixed signal is expected to reflect
the actual vegetation types in the region although it may be challenging to correctly assign
particular pollen types to vegetation units. For instance, Ivanov (2001) interpreted a pollen
diagram from a Tortonian section in the Sandanski Graben (Bulgaria; 8 in Fig. 1) with a
considerable amount of herbaceous pollen (including Poaceae, Chenopodiaceae, and
Artemisia making up c. 5 to 20% of the pollen spectrum) to reflect extra-regional open
vegetation on an elevated plateau in addition to swamp forests, riparian forests, and mixed
mesophytic forests developed in a river valley and adjoined slopes. Here, downslope
transportation of pollen from open landscapes blurred the local signal of the pollen record, but
at the same time added regional and extra-regional vegetation information.
A close relationship between the actual vegetation and the pollen spectrum from recent and
Holocene sediment samples has also been reported for NW Turkey (van Zeist & Bottema,
1991). Modern surface-sample spectra accurately depicted the regional vegetation, although
some taxa were underrepresented in the pollen spectra while others were overrepresented. For
example, pollen spectra of Fagus-Abies dominated areas showed relatively low percentages of
these two taxa (10.4% and 7.4%), while high amounts of Pinus (ca. 30%) derived from forests
thriving at some distance from the pollen trap. Likewise, comparatively high amounts of
Juniperus, Quercus, and Carpinus did not reflect the local vegetation but a regional signal. In
combination with a weak herbaceous signal (Poaceae < 5%; Chenopodiaceae,
Caryophyllaceae, Apiaceae represented by single pollen grains) the strong arboreal signal
provided a fairly accurate picture of the forest communities at a regional scale (van Zeist &
Bottema, 1991). In cases of bad pollen preservation (oxidised sedimentary rocks) it should be
kept in mind that only pollen with durable exines (high sporopollenin content) will be
preserved (e.g. Pinus, Chenopodiaceae/Amaranthaceae; Traverse, 2007) resulting in a biased
signal.
In contrast to dispersed spores and pollen, macrofossils (leaves) mainly reflect local and
regional vegetation whereas extra-regional vegetation is usually not reflected. Leaf remains
are mostly scattered isolated carbonised compression fossils, which are not concentrated
abundantly (“Blätterton” layers or paper shales) in distinct fossiliferous layers (Kvaček et al.,
2002); both small and larger leaves are usually not fragmented and hence there is no
indication for long-distance transport in high-energy depositional settings. At the same time,
low pollen abundances of evergreen oaks along with abundant leaf fossils representing
10
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
evergreen oaks might indicate that these leaves were transported by slow-flowing streams
over relatively large distances. Rarely, large fruit bodies are encountered, mostly represented
by conifer cones. Therefore, a combined wind and water transport from habitats bordering the
lake can be assumed (Kvaček et al., 2002). Among woody plants, Fagaceae (Fagus, Quercus)
are ecologically diverse and niche conserved at the genus/section level. Fagus is exclusively
found on well-drained soils and hence was not an element of the swamp forest vegetation.
However, under humid equable climates, lowland coastal and deltaic forests may contain
Fagus and hardwood hammocks with rich broadleaf deciduous and evergreen forests may be
present next to aquatic and hydric vegetation (Ferguson et al., 1999). In contrast, white oaks,
sect. Quercus, may thrive in swamp forests, riparian forests, mesic forests of lowlands and
uplands, or may form Mediterranean scrub. These different ecologies are well reflected in leaf
morphology whereas pollen morphology at the sectional level does not discriminate different
species/ecologies (Denk & Grimm, 2009). Since white oaks are represented by pollen only,
no further conclusions can be drawn as to their ecologies. Other sections of Quercus (sects
Cerris, Ilex) represented in the Vegora assemblage are highly niche conserved and exclusively
found on well-drained soils. Based on differences in leaf morphology (leaf size,
deciduousness), fossil-taxa such as Quercus gigas (leaf lamina up to 22 cm long; sect. Cerris)
might indicate humid temperate conditions on northern slopes (Kvaček et al., 2002), while
Quercus kubinyi (Cerris) might have been part of drier slopes. These fossil-species could
have been accessory elements in Fagus-dominated or oak-dominated forests (see above,
Results section). Section Ilex comprises evergreen species exclusively growing on welldrained soils. Closest relatives of the Messinian taxa are modern Mediterranean species (the
fossil-species Q. sosnowskyi resembles the modern species Q. alnifolia, endemic to Cyprus,
by leaf shape and leaf epidermal features; Kvaček et al., 2002) and Himalayan/East Asian
species (e.g. Q. drymeja resembles the modern Q. floribunda, south of the Himalayas; Denk
et al., 2017). Inferring the ecological properties of these fossil-taxa is not straight-forward:
Morphologically they either resemble modern E Mediterranean taxa or temperate Himalayan
taxa. At the same time, time-calibrated molecular phylogenies suggest that the modern
Mediterranean members of sect. Ilex diverged from their Himalayan sister species during late
Oligocene to early Miocene times, long before the deposition of the plant assemblage of
Vegora (Jiang et al., 2019). Within the Mediterranean clade, the most mesic species Q. ilex
also occurs in humid temperate forests of the Euxinian region (N Turkey, W Georgia) and
diverged from the remaining species of western Eurasian sect. Ilex no later than 9 Ma (Jiang
et al., 2019). Assuming that fully Mediterranean climate conditions, with precipitation
minima during the summer, in the Mediterranean region did not establish prior to the early
Pliocene (Suc, 1984; Velitzelos et al., 2014) we speculate that the Messinian members of sect.
Ilex were chiefly temperate species that went extinct during the Pliocene (cf. Jiang et al.,
2019). Specifically, Q. drymeja might have formed a forest belt above the Fagetum
gussonii/Quercetum mixtum and below the needleleaf evergreen forest belt above. Other
Quercus sect. Ilex such as Q. mediterranea and Q. sosnowskyi may have formed woody
shrublands or forests on drier sites (edaphically or due to the aspect of the slope).
Concerning the presence of grasslands or open woodlands the palaeobotanical data at hand
cannot discriminate between different scenarios. For taxa that are known from the macrofossil
record (Chamaerops, evergreen oaks) it is almost certain that they were part of the regional
flora of the FPS. The woody genera Olea, Cotinus, and Pistacia, known only from the pollen
record of unit 2, are typical elements of the present Mediterranean and submediterranean
vegetation belt in S Europe. Bell & Fletcher (2016) found that soil samples in open vegetation
11
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
plots in N Morocco recorded 20% to 35% arboreal pollen (AP). Main contributors to this
regional to extra-regional airborne pollen rain were Quercus types and Olea. In our sample,
arboreal pollen makes up almost 90% of the total count. Quercus certainly was a major
component of local to regional vegetation because it is the most prominent component with
several deciduous and evergreen species in the leaf flora of Vegora. In contrast, Olea makes
up less than 1% in the pollen count. No leaf and seed remains reminiscent of Olea are
recorded from Vegora. This along with the known tendency of/ability for long-distance
transport (Bell & Fletcher, 2016) might indicate the presence of Olea at a greater distance
from the Vegora basin. The same can be assumed for Cotinus and Pistacia. The latter,
however, do also occur in open-canopy pine forests.
In case of herbaceous taxa represented by single or very few pollen grains in the palynological
record, these may also reflect long-distance dispersal from high mountain or even from more
distant coastal areas to the west of the FPS. They would then provide an extra-regional
vegetation signal. Potential elements of open vegetation include Apiaceae, Chenopodiaceae,
Poaceae, Geranium, Caryophyllaceae, and Cichorioideae. Except for Poaceae and
Chenopodiaceae these taxa are predominantly insect-pollinated. For wind-pollinated taxa
represented with 1–3 grains in the pollen count (Poaceae, Chenopodiaceae) we assume that
this is indicative of a regional or extra-regional source vegetation. The insect-pollinated taxa,
also represented by 1–3 grains in the pollen count, are difficult to assign to either local or
regional/extra-regional vegetation. If these groups were local elements, they would have been
quite rare, based on the low numbers of their pollen. They could have been part of the
lakeshore vegetation, of open rocky places, of the understorey of forest vegetation or
meadows above the tree line. Alternatively, these elements could have been brought in by
long-distance dispersal (LDD) from coastal plains to the southeast and east of the FPS.
In sum, the combined macrofossil and microfossil record offers an accurate picture of the
different vegetation types present in the FPS during the Messinian. The fossil record suggests
that the local and regional vegetation in the FPS comprised a range of ecologically different
zonal and azonal forest types, while LDD of several herbaceous taxa may potentially have
contributed to an extra-regional pollen signal.
4.2. Inferring Messinian pre-evaporitic vegetation of the FPS and adjacent areas
Our multi-proxy palaeobotanical study of the Messinian assemblage of Vegora is based on
information from fruits and seeds, leaves, and dispersed pollen and spores. For the main flora
in unit 2 (blue marls) we used information from leaf fossils and dispersed spores and pollen.
As discussed above, there is strong evidence for the presence of a wide range of forest and
forest/shrubland types in the FSP. Furthermore, a small number of woody and herbaceous
taxa could reflect open vegetation. The latter are represented by low numbers of pollen grains,
which could be ascribed to LDD from remote areas including dry uplands or coastal plains. In
order to evaluate the pre-evaporitic Messinian vegetation of the FPS, we compared our finds
with previously published data on other plant fossil localities in the FPS and surrounding
areas. In addition, a vertebrate locality from the Axios valley (Dytiko1, 2, 3; 3 in Fig. 1) is
roughly coeval with the Messinian pre-evaporitic assemblages of the FPS (Koufos, 2006).
The hypsodonty index of this fauna of 1.45-1.86 (NOW database,
http://pantodon.science.helsinki.fi/now/locality.php?p=ecometrics) corresponds to the diet
types “mixed-closed habitats”, “regular browsers”, and “selective browsers” according to
Janis (1988) and hence provides an excellent match with the environments inferred for the
FPS.
12
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
From Lava (Fig. 1), Steenbrink et al. (2000) investigated two sequences covering two
sedimentary cycles each. Based on palaeomagnetic correlation these sequences are dated as c.
6.8–6.7 Ma and c. 6.3 Ma. The pollen assemblages are comparable to the Vegora assemblage
but differ in some respects. First, the Lava sections have a continuous high amount of Pinus
pollen (20 to >60%) suggesting that the fossil site was located closer to pine forests than was
the Vegora lake. Second, Cedrus pollen is abundant with values between 10 and >30%. Third,
Steenbrink et al. (2000) did not report evergreen oak pollen, although evergreen Quercus is
known from Lava based on leaf fossils (Velitzelos et al., 2014). Steenbrink et al. (2000)
inferred a humid temperate climate without dry season for the investigated sedimentary
cycles. In addition, they suggested that expansions of Fagus accompanied by a decrease of
Abies might reflect subtle increases in montane humidity. Overall, they suggested
continuously wet and warm-temperate climate conditions for the investigated period for Lava.
Velitzelos et al. (2014) provided revised taxon lists for the roughly coeval macrofossil (leaves
and fruits/seeds) localities Prosilio and Lava (age based on palaeomagnetic correlation, 6.7–
6.4 Ma; Steenbrink et al., 2006). The macroflora is very similar to the one from Vegora in
terms of composition. However, whereas Quercus sosnowskyi is among the most abundant
elements in Vegora, only a few leaves represent this species in Prosilio; also Glyptostrobus is
much less abundant. Pinus is represented by cones, leaf fascicles, and leafy branches; this is
in accordance with the high amount of pine pollen documented in the palynological record.
Fagus is a frequent element as well, while Abies is not recorded in the macroflora.
Likoudi, 20 km S of Lava, is located in a small basin south of the main FPS (Fig. 1). The
macroflora (leaves, fruits and seeds) is very rich (see revised and updated floral list in
Velitzelos et al., 2014). The precise age of the Messinian diatomaceous marls is not clear
(Knobloch & Velitzelos, 1986) although it unambiguously is pre-evaporitic. The flora is
characterised by the high diversity of conifers (11 genera of Cupressaceae and Pinaceae,
including Torreya – as Egeria sp. in Velitzelos et al., 2014). As in Vegora, Fagus is a
dominating element. Other taxa (Cercis, Laria, cf. Nerium) are not known from other FBS
floras. Well-preserved cones of Cedrus and cones and leafy twigs of Cathaya and Taiwania
suggest that these genera were not growing at high elevations but nearby the area of
deposition (lake). If coeval with the Lava deposits, this would explain the relatively high
amounts of Cedrus pollen in the palynological section of Lava.
Ivanov & Slavomirova (2002) investigated a 70 m succession of lacustrine sediments in the
Bitola Basin (Northern Macedonia; borehole V-466; 1 in Fig. 1) about 10 km E of Bitola and
40 km NNW of Vegora. Based on a vertebrate fauna on top of these sediments the plantbearing sediments are assigned a late Miocene age (Dumurdžanov et al., 2002; OgnjanovaRumenova, 2005). From these sediments, abundant leaves of Quercus sosnowskyi have been
reported (Dumurdžanov et al., 2002). The pollen assemblage is similar to the one from
Vegora by its high amounts of woody taxa (both conifers and angiosperms) and the
composition and very low abundance of herbaceous taxa. In contrast to the Vegora
assemblage, few taxa are represented by high percentages (Pinus 20–40%, Fagus up to 10%,
Quercus 10–20%, Taxodioideae and Sequoioideae 10–15%, and Alnus, up to 20%). In the
Vegora sample, only Pinus reached more than 10% in the pollen count.
For N and C Italy, Kovar-Eder et al. (2006) and Bertini & Martinetto (2008) reported
prevalence of deciduous Fagus and Quercus in the northern parts whereas in addition
sclerophyllous plants (e.g. Quercus mediterranea) were common in Messinian pre-evaporitic
assemblages of C and S Italy (Gabbro I, Senigallia, Palena).
13
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
From the Serres Basin (7 in Fig. 1), Psilovikos & Karistineos (1986), Karistineos & Ioakim
(1989) and Suc et al. (2015) reported pollen assemblages that reflect lowland swamp forests
(NLD and BLD) and hinterland vegetation including broadleaf deciduous and needleleaf
evergreen forests. The age of these localities is younger than Vegora and based on mammal
data (Koufos, 2006) and geological data (Suc et al., 2015) might have been deposited during
the post-evaporitic Messinian.
It is remarkable that post-evaporitic and evaporitic Messinian floras from Italy (Bertini &
Martinetto, 2008) and Greece (Velitzelos et al., 2014) mainly reflect moist conditions and
persistence of forested environments. This is in strong contrast to conditions from Turkey,
where steppe and steppe forest were established by the end of the middle Miocene (Denk et
al., 2018).
Main vegetation types recognised for the pre-evaporitic Messinian of the FPS and their
modern analogues are summarised in Table 3.
4.3. Inferring Messinian pre-evaporitic climate of the FPS and adjacent areas
Fruit/seed and leaf records mainly represent azonal vegetation and may not adequately reflect
the zonal vegetation growing at some distance from the area of sedimentation. Zonal
vegetation is mainly controlled by large-scale climate (regional, extra-regional). Azonal
vegetation is controlled by edaphic conditions rather than large-scale climate. For these
reasons, seed/fruit and leaf assemblages either used for physiognomic (CLAMP) or
taxonomic (Köppen signatures, CA) methods may not be suitable for a meaningful
(representative) climate reconstruction or they may reconstruct a local climate. However, as
shown by Ferguson et al. (1999), under mild and humid climate conditions, the vegetation
close to the area of sedimentation consists of a mosaic of vegetation types, some of which are
composed of woody plants that also dominate the hinterland vegetation. For example, natural
levee and hardwood hammock vegetation associated with azonal swamp, riparian and bog
vegetation in SE North America contains taxa such as Fagus, Magnolia, Ilex spp., Carpinus,
and Symplocos. In this case, the limitations outlined above are not valid. In drier, strongly
seasonal climate settings, the azonal vegetation will not be sufficient to produce meaningful
(representative) regional climate estimates.
When the palynological record is taken into consideration, both the local and the regional
(hinterland, vertical vegetation belts) vegetation is likely to be captured. This is reflected in
the much greater diversity of conifers and herbaceous taxa in the pollen record as compared to
the macrofossil record in the Vegora assemblage. In addition, long-distance dispersal may add
information from the extra-regional vegetation (e.g. Olea; van Zeist & Bottema, 1991; Bell &
Fletcher, 2016).
The resulting differences in reconstructed climate/vegetation are seen in the Köppen
signatures for the leaf fossil and pollen/spore records of Vegora (Fig. 7). Taxa extending into
tropical climates are much better represented in the pollen record than in the leaf record.
Therefore, Köppen signatures, although not providing exact values for different climate
parameters, offer important qualitative information. This can be illustrated by comparing
three floras from the E Mediterranean region, all dominated by Quercus spp. and Fagus and
ranging in age from early Burdigalian to Messinian: The early Burdigalian flora of Güvem,
Anatolia (MN3, 20–18 Ma; Denk et al., 2017) has a tropical signal of 16% in Köppen
signatures from pollen and spore assemblages, whereas tropical signal from leaf fossils is
11.9%. The middle Miocene floras of the Yatağan basin (Langhian/Serravallian,
MN6/MN7+8; 14.8–13.8 Ma; Bouchal et al., 2018) show a trend from the more humid MN6
14
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
zone to the more arid MN7+8 zone with tropical signal from pollen and spore assemblages
decreasing from 11% in pollen zones 1 and 2 to 7% in the transitional pollen zone 2/3. The
tropical signal in Köppen signatures from the macrofossil assemblage from pollen zone 2 is
9%. In the pre-evaporitic assemblages of Vegora, the tropical signal is 9% and 6% for the
pollen/spore and leaf assemblages.
In contrast, the coexistence approach (CA), in producing exact values for selected climate
parameters, will be prone to produce hybrid climates if fossil assemblages represent a range
of lowland azonal and hinterland vegetation and vegetation from different vertical belts. Such
artificial climates may be randomly expressed as narrow coexistence intervals including
different vertical vegetation belts, or in very broad coexistence intervals (cf. Table 2). Thus,
Köppen signatures are considered a more dynamic way of climate reconstruction as they
account for the possibility that different elements may derive from different vertical
vegetation belts.
CLAMP (Physg3arcAZ) reconstructed WMMT that corresponds with a temperate climate
with warm summers, Cfb Köppen climate type, underscoring the temperate character of the
Vegora assemblage. The calibration set Asia 1 reconstructed a cooler climate with CMMT
below 0°C. This would appear to be in conflict with the cold tolerance of the palm
Chamaerops that is recorded from the leaf record and palms in general as recorded in the
pollen record. However, all three approaches to reconstruct the pre-evaporitic Messinian
climate of the FPS suggest a temperate climate with weak seasonality (Table 2).
In view of the relatively homogeneous vegetation signal from the FPS (both from dispersed
spores/pollen and leaves), the inferred climate is highly plausible.
5. Conclusion
The present study used palaeobotanical data to reconstruct palaeoenvironments in an
intermontane basin of NW Greece shortly before the Messinian salinity crisis. For the period
6.4-6 Ma, leaf fossil data and dispersed spores and pollen indicate the presence of various
types of forest including riparian and mesic forests, deciduous and evergreen forests,
laurophyllous forests and sclerophyllous woodlands. Open landscapes dominated by
herbaceous plants were not reconstructed for the Florina-Ptolemais-Servia Basin. Based on
sparse pollen records resulting from potential long-distance dispersal it cannot be ruled out
that herbaceous plants played a more important role in coastal lowlands to the W of the study
area. However, contemporaneous vertebrate faunas from the Axios valley also suggest mixed
forested vegetation. We found that the combined leaf and dispersed spore/pollen records
allow fairly accurate reconstruction of local and regional vegetation. Leaf fossils offer more
species-diagnostic features than pollen and a combination of leaf taphonomy and pollen
frequencies allow discriminating local and regional vegetation. Furthermore, specific
comparison with modern pollen spectra was made in order to understand biases in pollen
abundances. This provided a transfer function for the interpretation of fossil pollen
assemblages. The results of this study confirm previous findings of a N to S gradient of
temperature and precipitation seasonality in the Mediterranean area. Our results also reinforce
the notion that steppe and forest steppe environments evolved earlier in Turkey with
deciduous oaks playing important roles in the woody flora. In contrast, laurophyllous
evergreen oak forest persisted in Greece/Italy into the late Miocene and as relict into the
Pliocene. Climate reconstructions utilizing three different approaches to climate
reconstruction resulted in roughly similar values that translate into a cool temperate Cfb
climate according to the Köppen-Geiger climate classification. When comparing CLAMP,
15
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
CA and Köppen signatures, we noticed that leaf assemblages may be biased towards
autochthonous plant communities and by this the climate signal will be highly local.
However, under humid mild climates plants from the hinterland vegetation may thrive from
lowlands to high altitudes and interspersed in riparian landscapes (e.g. on hammocks, gallery
forests etc.). Furthermore, the taphonomy of the fossil plant assemblage of Vegora suggests
that slow-flowing streams had transported leaves from different vegetation types of the
hinterland into the lake and thus the leaf taphocoenoses would be representative of both the
local and regional vegetation. This causes a further problem for CLAMP and more so for CA:
climate signals in the fossil plant assemblage may be highly mixed and may derive from
different vertical vegetation belts. Köppen signatures, while not generating exact values for
particular climate parameters, overcome the problem of hybrid climates as they collect
climatic/environmental signal in a plant assemblage without averaging different signals into a
single, possibly artificial, signal.
Data accessibility. All data used in this article are made available in the Supplemental
Material.
Author’s contributions. T.D. designed the study and wrote the first draft. J.M.B.
investigated dispersed pollen and spores and made the Köppen signatures and CA analyses.
T.H.G. made the CLAMP analysis. T.D., J.M.B., T.H.G., D.V., and E.V. evaluated the fossil
data and wrote the final paper.
Competing interests. We declare we have no competing interests.
Funding. This study was financed by a grant of the Swedish Research Council (VR; project
no. 2015-03986) to T.D. T.H.G. was supported by The Scientific and Technological Research
Council of Turkey (TÜB İTAK), 2219 Post-Doctoral Research Fellowship Program (2017/2),
project no. 1059B191700382.
Supplemental Material. The supplement related to this article is available online at:
https://figshare.com/articles/JMB_THG_DV_EV_TD_Supplementary_Material/10327646
References
Akhani H, Djamali M, Ghorbanalizadeh A, Ramezani E. 2010 Plant biodiversity of Hyrcanian
relict forests, N Iran: an overview of the flora, vegetation, palaeoecology and
conservation. Pak. J. Bot. 42, 231–258.
Akkemik Ü. 2003 Tree rings of Cedrus libani at the northern boundary of its natural
distribution. IAWA J. 24, 63–73.
Angiosperm Phylogeny Group (APG). 2016 An updated of the angiosperm phylogeny group
classification for the orders and families of flowering plants APG IV. Bot. J. Linn. Soc.
181, 1–20.
Bell BA, Fletcher WJ. 2016 Modern surface pollen assemblages from the Middle and High
Atlas, Morocco: insights into pollen representation and transport. Grana 55, 286–301.
Bertini A, Martinetto E. 2008 Messinian to Zanclean vegetation and climate of Northern and
Central Italy. Boll. Soc. Paleontol. Ital. 47, 105–121.
Brunn JH. 1956 Contribution à l’étude géologique du Pinde septentrional et d’une partie de la
Macédonie. Ann. Géol. Pays Hellèn. 7, 1–358.
Bouchal JM, Güner TH, Denk T. 2018 Middle Miocene climate of southwestern Anatolia
from multiple botanical proxies. Clim. Past 12, 1427–1440.
Budantsev LY, Golovneva LB. 2009 Fossil Flora of Arctic II - Palaeogene Flora of
Spitsbergen. St. Petersburg: Komarov Bot. Inst. Russ. Acad. Sci.
16
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
Cerling TE, Harris JM, MacFadden BJ, Leakey MG, Quade J, Eisenmann V, Ehleringer JR.
1997 Global change through the Miocene/Pliocene boundary. Nature 389, 153–158.
Connor SE, Thomas I, Kvavadze EV, Arabuli GJ, Avakov GS, Antonio Sagona A. 2004 A
survey of modern pollen and vegetation along an altitudinal transect in southern
Georgia, Caucasus region. Rev. Palaeobot. Palyn. 129, 229–250.
Daghlian CP. 1982 A simple method for combined light, scanning and transmission electron
microscope observation of single pollen grains from dispersed pollen samples. Pollen et
Spores 24, 537–545.
Denk T, Grimm GW. 2009 Significance of pollen characteristics for infrageneric
classification and phylogeny in Quercus (Fagaceae). Int. J. Plant Sci. 170, 926–940.
Denk T, Velitzelos D. 2002 First evidence of epidermal structure of Ginkgo from the
Mediterranean Tertiary. Rev. Palaeobot. Palynol. 120, 1–15.
Denk T, Frotzler N, Davitashvili N. 2001 Vegetational patterns and distribution of relict taxa
in humid temperate forests and wetlands of Georgia (Transcaucasia). Biol. J. Linn. Soc.
72, 287–332.
Denk T, Grimm GW, Grímsson F, Zetter R. 2013 Evidence from “Köppen signatures” of
fossil plant assemblages for effective heat transport of Gulf Stream to subarctic North
Atlantic during Miocene cooling, Biogeosci. 10, 7927–7942. https://doi.org/10.5194/bg10-7927-2013.
Denk T, Güner TH, Kvaček Z, Bouchal MJ. 2017 The early Miocene flora of Güvem (Central
Anatolia, Turkey): a window into early Neogene vegetation and environments in the
Eastern Mediterranean. Acta Palaeobot. 57, 237–338.
Denk T, Velitzelos E, Güner HT, Bouchal MJ, Grímsson F, Grimm GW. 2017 Taxonomy and
palaeoecology of two widespread western Eurasian Neogene sclerophyllous oak
species: Quercus drymeja Unger and Q. mediterranea Unger. Rev. Palaeobot. Palynol.
241, 98–128.
Denk T, Zohner CM, Grimm GW, Renner SS. 2018 Plant fossils reveal major biomes
occupied by the late Miocene Old-World Pikermian fauna. Nature Ecol. Evol. 2, 1864–
1870.
Denk T, Güner HT, Bouchal JM. 2019 Early Miocene climate and biomes of Turkey:
Evidence from leaf fossils, dispersed pollen, and petrified wood. Palaeogeogr.,
Palaeoecol., Palaeoclimatol. 530, 236–248.
Dolezych M, Schneider W. 2006 Inkohlte Hölzer und Cuticulae dispersae aus dem 2.
Miozänen Flözhorizont im Tagebau Welzow (Lausitz) – Taxonomie und vergleichende
feinstratigraphisch-fazielle Zuordnung. Z. geol. Wiss. 34, 165–259.
Dolezych M, Schneider W. 2007 Taxonomie und Taphonomie von Koniferenhölzern und
Cuticulae dispersae im 2. Lausitzer Flözhorizont (Miozän) des Senftenberger Reviers.
Palaeontograph. B 276, 1–95.
Dolezych M, Schneider W. 2012 Fossil conifers in peat bog environments – results from the
Central European Neogene. Japan. J. Palynol. 58 (Special Issue), 48.
Dumurdžanov N, Krstić N, Mihajlović D, Ognjanovo-Rumenova N, Petrov G. 2002 New data
on stratigraphy of the Neogene and Pleistocene in Mariovo, Macedonia. Geol.
Macedon. 17, 43–52.
Erdős L, Ambarli D, Anenkhonov OA, Bátori Z, CserhalmiD, Kiss M, Kröel-Dulay G, Liu H,
Magnes M, Molnár Z, et al. 2018 The edge of two worlds: A new synthesis on Eurasian
forest-steppes. Appl. Vegetation Sci. 21, 345–362.
17
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
Favre E, Escarguel G, Suc J-P, Vidal G, Thévenod L. 2008 A contribution to deciphering the
meaning of AP/NAP with respect to vegetation cover. Rev. Palaeobot. Palynol. 148,
13–35.
Fang J, Wang Z, Tang Z. 2009 Atlas of Woody Plants in China. Volumes 1 to 3 and index.
Beijing: Higher Education Press.
Fauquette S, Suc J-P, Bertini A, Popescu S-M, Warny S, Bachiri Taoufiq N, Perez Villa, M-J,
Chikhi H, Subally D, Feddi N, Clauzon G, Ferrier J. 2006 How much did climate force
the Messinian salinity crisis? Quantified climatic conditions from pollen records in the
Mediterranean region. Palaeogeogr., Palaeoclimatol.,Palaeoecol. 238, 281–301.
Ferguson DK, Hofmann C-C, Denk T. 1999 Taphonomy: field techniques in modem
environments. In: Jones TP, Rowe NP (eds) Fossil Plants and Spores: modern
techniques. London: Geological Society, pp 210-213.
Freitag, H. 1971 Die natürliche Vegetation Afghanistans. Vegetatio 22, 285–344.
Gersonde R, Velitzelos E. 1978 Diatomeenpaläoökologie im Neogen-becken von Vegora NW Mazedonien (vorläufige Mitteilung). Ann. Gèol. Pays Hellèn. 30, 373–382.
Grimm GW, Bouchal JM, Denk T, Potts A. 2016 Fables and foibles: A critical analysis of the
Palaeoflora database and the Coexistence Approach for palaeoclimate reconstruction.
Rev. Palaeobot. Palyn. 233, 611–622.
Grimm GW, Potts A. 2016 Fallacies and fantasies: the theoretical underpinnings of the
Coexistence Approach for palaeoclimate reconstruction. Clim. Past 12, 611–622.
Halbritter H, Ulrich S, Grímsson F, Weber M, Zetter R, Hesse M, Buchner R, Svojtka M,
Frosch-Radivo A. 2018 Illustrated Pollen Terminology, 2nd ed. Cham (Switzerland):
Springer Nature.
Herbert TD, Lawrence KT, Tzanova A, Peterson LC, Caballiero-Gill R, Kelly CS. 2016 Late
Miocene global cooling and the rise of modern ecosystems. Nature Geosci. 9, 843–847.
Ivanov DA. 2001 Palaeoecological interpretation of a pollen diagram from the Sandanski
graben (Southwest Bulgaria). Comptes Rend. Acad. Bulg. Sci. 54, 65–68.
Ivanov DA, Slavomirova E. 2002 Preliminary palynological data on Neogene flora from
Bitola Basin (F.Y.R.O.M.). Comptes Rend. Acad. Bulg. Sci. 55, 81–86.
Janis CM. 1986 An estimation of tooth volume and hypsodonty indices in ungulate mammals,
and the correlation of these factors with dietary preference. In: DE Russell, JP Santoro,
D Sigogneau-Russell (eds) Teeth revisited. Proc. 7th Int. Symp. Dental Morph. Mém.
Mus. Nat. Hist. naturelle, Sér. C 53, 367–387.
Jiang X-L, Hipp AL, Deng M, Su T, Zhou Z-K, Yan M-X. 2019 East Asian origins of
European holly oaks (Quercus section Ilex Loudon) via the Tibet-Himalaya. J.
Biogeogr. 46, 2188–2202. doi.org/10.1111/jbi.13654
Karistineos N, Ioakim C. 1989 Palaeoenvironmental and palaeoclimatic evolution of the
Serres Basin (N. Greece) during the Miocene. Palaeogeogr., Palaeoclimatol.,
Palaeoecol. 70, 275–285.
Knobloch E, Velitzelos E. 1986a Die obermiozäne Flora von Likudi bei Elassona (Thessalien,
Griechenland). Doc. Nat. 29, 5–20.
Knobloch E, Velitzelos E. 1986b Die obermiozäne Flora von Prosilion bei Kozani (SudMazedonien, Griechenland). Doc. Nat. 29, 29–33.
Kottek M, Grieser J, Beck C, Rudolf B, Rubel F. 2006 World map of the Köppen-Geiger
climate classification updated. Meteorol. Z. 15, 259–263.
Koufos GD. 1982 Hipparion crassum Gervais, 1859 from the lignites of Ptolemais
(Macedonia-Greece). Proc. Kon. Ned. Akad. Weten. B. 85, 229–239.
18
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
Koufos GD. 2006 The Neogene mammal localities of Greece: Faunas, chronology and
biostratigraphy. Hellen. J. Geosci. 41, 183–214
Koufos GD, Kostopoulos DS, Koliadimou KK. 1991 Un nouveau gisement de mammiferes
dans le Villafranchien de Macedoine occidentale (Gréce). C. R. Acad. Sci. Paris, ser. II
313, 831–836.
Kovar-Eder J, Jechorek H, Kvaček Z, Parashiv V. 2008 The Integrated Plant Record: An
essential tool for reconstructing Neogene zonal vegetation in Europe. Palaios 23, 97–
111.
Kovar-Eder J, Kvaček Z, Martinetto E, Roiron P. 2006 Late Miocene to Early Pliocene
vegetation of southern Europe (7-4 Ma) as reflected in the megafossil plant record.
Palaeogeogr., Palaeoclimatol., Palaeoecol. 238, 321–339.
Kozlowski G, Bétrisey S, Song Y. 2018 Wingnuts (Pterocarya) and walnut family. Relict
trees: linking the past, present and future. Natural History Museum Fribourg:
Switzerland.
Kvaček Z, Velitzelos D, Velitzelos E. 2002 Late Miocene Flora of Vegora Macedonia N.
Greece. Athens: Koralis.
Maharramova E. 2015 Genetic diversity and population structure of the relict forest trees
Zelkova carpinifolia (Ulmaceae) and Pterocarya fraxinifolia (Juglandaceae) in the
South Caucasus. PhD Dissertation. Freie Universität Berlin.
Mai HD. 1995 Tertiäre Vegetationsgeschichte Europas. Stuttgart: Gustav Fischer.
Mai DH, Velitzelos E. 1992 Über fossile Pinaceen-Reste im Jungtertiär von Griechenland.
Feddes Repert. 103, 1–18.
Mai DH, Velitzelos E. 1997 Paläokarpologische Beiträge zur jungtertiären Flora von Vegora
(Nordgriechenland). Feddes Repert. 108, 507–526.
Marinova E, Harrison SP, Bragg F, Connor S, de Laet V, Leroy SAG, Mudie P, Atanassova J,
Bozilova E, Caner H, Cordova C, Djamali M, Filipova-Marinova M, Gerasimenko N,
Jahns S, Kouli K, Kotthoff U, Kvavadze E, Lazarova M, Novenko E, Ramezani E,
Röpke A, Shumilovskikh L, Tanţău I, Tonkov S. 2018 Pollen-derived biomes in the
Eastern Mediterranean–Black Sea–Caspian-Corridor. J. Biogeogr. 45, 484–499.
Mayer H, Aksoy, H. 1986 Wälder der Türkei. Stuttgart: Gustav Fischer.
Mosbrugger V, Utescher T. 1997 The coexistence approach — a method for quantitative
reconstructions of Tertiary terrestrial palaeoclimate data using plant fossils.
Palaeogeogr., Palaeoclimatol., Palaeoecol. 134, 61–86
Ognjanova-Rumenova N. 2005 Upper Neogene siliceous microfossils from Pelagonia Basin
(Balkan Peninsula). Geol. Carpathica 56 (4), 347–358.
Pavlides SB, Mountrakis DM. 1986 Neotectonics of the Florina–Vegoritis–Ptolemais
Neogene Basin (NW Greece): an example of extensional tectonics of the greater Aegean
area. Ann. Géol. Pays Hellen. 33, 311–327.
Peel MC, Finlayson BL, McMahon TA. 2007 Updated world map of the Köppen-Geiger
climate classification. Hydrol. Earth Syst. Sci., 11, 1633–1644.
Psilovikos A, Karistineos N. 1986 A depositional sedimentary model for the Neogene
uraniferous lignites of the Serres graben, Greece. Palaeogeogr., Palaeoclimatol.,
Palaeoecol. 56, 1–16.
Punt W, Hoen PP, Blackmore S, Nilsson RH, Le Thomas A. 2007 Glossary of pollen and
spore terminology. Rev. Palaeobot. Palynol. 143, 1–81.
19
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
Ratnam J, Bond WJ, Fensham RJ, Hoffmann WA, Archibald S, Lehmann CER, Anderson
MT, Higgins SI, Sankaran M. 2011 When is a ‘forest’ a savanna, and why does it
matter? Glob. Ecol. Biogeogr. 20, 653–660.
Roberts S, Rassios A, Wright L, Vacondios I, Vrachatis G, GrivasE, Nesbitt RW, Neary CR,
Moat T, Kostantopolou L. 1988 Structural controls on the location and form of the
Vourinos chromite deposits. In: Boissonnas J, Omenetto P (eds) Mineral deposits within
the European Community. Springer, Berlin Heidelberg New York, pp 249–266.
Rubel F, Brugger K, Haslinger K, Auer I. 2017 The climate of the European Alps: Shift of
very high resolution Köppen-Geiger climate zones 1800–2100. Meteorol. Z., 26, 115–
125. https://doi.org/10.1127/metz/2016/0816.
Schneider W. 1992 Floral successions in the Miocene swamps and bogs of Central. Europe.
Z. Geol.Wiss. 20, 555–570.
Spicer RA. 2008 CLAMP. In: Gornitz V. Ed. Encyclopedia of Paleoclimatology and Ancient
Environments. Dordrecht: Springer.
Steenbrink J, Van Vugt N. Hilgen FJ, Wijbrans JR, Meulenkamp JE. 1999 Sedimentary
cycles and volcanic ash beds in the lower Pliocene lacustrine succession of Ptolemais
(NW Greece): discrepancy between 40Ar/39Ar and astronomical ages. Palaeogeogr.,
Palaeoclimatol., Palaeoecol. 152, 283–303.
Steenbrink J, Van Vugt N, Kloosterboer-van Hoeve ML, Hilgen FJ. 2000 Refinement of the
Messinian APTS from sedimentary cycle patterns in the lacustrine Lava section (Servia
Basin, NW Greece). Earth Planet. Sci. Lett. 181, 161–173.
Steenbrink J, Hilgen FJ, Krijgsman W, Wijbrans JR, Meulenkamp JE. 2006 Late Miocene to
Early Pliocene depositional history of the intramontane Florina-Ptolemais-Servia Basin,
NW Greece: Interplay between orbital forcing and tectonics. Palaeogeogr.,
Palaeoclimatol., Palaeoecol. 238, 151–178.
Suc J-P. 1984 Origin and evolution of the Mediterranean vegetation and climate in Europe.
Nature 307, 429–432.
Suc J-P, Popescu S-M, Do Couto D, Clauzon G, Rubino J-L, Melinte-Dobrinescu MC,
Quillévéré F, Brun J-P, Dumurdžanov N, Zagorchev I, Lesdić V, Tomić D, Sokoutis D,
Meyer B, Macaleţ R, Jelen B, Rihelj H. 2015 Marine gateway vs. fluvial stream within
the Balkans from 6 to 5 Ma. Mar. Pet. Geol. 66, 231–245.
Suc J-P, Popescu SM, Fauquette S, Bessedik M, Jiménez-Moreno G, Taoufiq B, Zheng Z,
Medail F, Klotz S. 2018 Reconstruction of Mediterranean flora, vegetation and climate
for the last 23 million years based on an extensive pollen dataset. Ecol. mediterr. 44,
53–85.
Traverse A. 2007 Paleopalynology. Topics in Geobiology 28. Dordrecht, Springer.
Utescher T, Bruch AA, Erdei B, François I, Ivanov D, Jacques FMB, Kern AK, Liu Y-SC,
Mosbrugger V, Spicer RA. 2014 The Coexistence Approach—Theoretical background
and practical considerations of using plant fossils for climate quantification.
Palaeogeogr., Palaeoclimatol., Palaeoecol. 410, 58–73.
Van de Weerd A. 1979 Early Ruscinian rodents and lagomorphs (Mammalia) from the
lignites near Ptolemais (Macedonia, Greece). Proc. Kon. Nederl. Akad. Wet., B. 82,
127–170.
Van Zeist W, Bottema S. 1991 Late Quaternary vegetation of the Near East. Beih. Tübinger
Atlas Vord. Orient, Reihe A (Naturwiss.) 18, 1–156.
Van Zeist W, Woldring H, Stapert, D. 1975 Late Quaternary vegetation and climate of
southwestern Turkey. Palaeohist. 14, 35–143.
20
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
Velitzelos D, Denk T. 2002 Leaf epidermal characteristics of late Tertiary conifers from
Greece: taxonomic significance and limitations. 6th Europ. Paleobot. Palynol. Conf.
Athens, Greece, Abstracts, 183–184.
Velitzelos D, Bouchal JM, Denk T. 2014 Review of the Cenozoic floras of Greece. Rev.
Palaeobot. Palynol. 204, 1–15.
Velitzelos E, Gregor H-J. 1985 Neue paläofloristische Befunde im Neogen Griechenlands.
Doc. Nat. 25, 1–4.
Velitzelos E, Krach JE, Gregor H-J, Geissert F. 1983 Bolboschoenus vegorae – ein Vergleich
fossiler und rezenter Rhizomknollen der Strandbinse. Doc. Nat. 5, 1–57.
Velitzelos E, Kvaček Z. 1999 Review of the late Miocene flora of Vegora western
Macedonia, Greece. Acta Palaeobotanica, Suppl. 2 (Proceed. 5th EPPC) 419–427.
Velitzelos E, Petrescu I. 1981 Seltene pflanzliche Fossilien aus dem Braunkohlebecken von
Vegora. Ann. géol. Pays hellén. 30, 767–777.
Velitzelos E, Schneider HE. 1979 Jungtertiäre Pflanzenfunde aus dem Becken von Vegora in
West-Mazedonien. 3. Mitteilung: Eine Fächerpalme (Chamaerops humulis L.). Ann.
géol. Pays hellén. 29, 796–799.
Woodward FI, Lomas MR, Kelly CK. 2004 Global climate and the distribution of plant
biomes. Phil. Trans. R. Soc. Lond. B. 359, 1465–1476.
Yang J, Spicer RA, Spicer TEV, Li C-S. 2011 'CLAMP Online': a new web-based
palaeoclimate tool and its application to the terrestrial Paleogene and Neogene of North
America. Palaeobiodiv. Palaeoenviron. 91, 163–183.
Zetter R. 1989 Methodik und Bedeutung einer routinemäßigen kombinierten
lichtmikroskopischen und rasterelektronenmikroskopischen Untersuchung fossiler
Mikrofloren. Cour. Forschungsinst. Senck. 109, 41–50.
Table and Figure Captions
Table 1. Plant taxa recorded from unit 1 (lignite seam) and unit 2 (blue marls) of the Vegora
section.
Table 2. Estimated climate parameters for the pre-evaporitic Messinian of Vegora from two
CLAMP calibration datasets and from CA.
Table 3. Vegetation types recognised for the pre-evaporitic Messinian of the Florina–
Ptolemais–Servia Basin.
Figure 1. Fossil localities and lithological map of the Florina–Ptolemais–Servia Basin.
Map redrawn after Steenbrink et al. (1999, 2006), Ognjanova-Rumenova (2005), Ivanov
(2001) and Koufos (2006). Fossil localities: (1) Bitola Basin, Republic of North Macedonia,
PF. (2) Vegora Basin, MF and PF (3) Dytiko, VF. (4) Prosilio, MF. (5) Lava, MF. (6)
Likoudi, MF. (7) Serres Basin. (2–7) Greece. (8) Sandanski Graben, Bulgaria, PF.
Abbreviations: Plant macrofossils (MF), palynoflora (PF), vertebrate fossils (VF).
Figure 2. Lithology and polarity zones of the Vegora section (redrawn after Steenbrink et al.,
2006). Position of fossil bearing strata following Velitzelos and Schneider (1979) and Kvaček
et al. (2002).
Figure 3. Light microscopy (LM) and scanning electron microscopy (SEM) micrographs of
algae, fern and fern allies, and gymnosperm palynomorphs.
(a) Botryococcus sp. cf. B. kurzii. (b) Spirogyra sp. 1/ Ovoidites elongatus. (c) Spirogyra sp.
2/Cycloovoidites cyclus. (d–e) Osmunda sp., (d) EV, (e) PV. (f) Cryptogramma vel
Cheilanthes sp, PV. (g–h) Pteris sp., (g) PV, (h) DV. (i) Davalliaceae vel Polypodiaceae sp./
21
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
Verrucatosporites alienus (R.Potonié) P.W.Thomson et Pflug, 1953, EV. (j)
Leavigatosporites haardti, EV. (k–l) Inaperturopollenites hiatus. (m) Abies sp., EV. (n–o)
Cathaya sp., (n) PV, (o) SEM detail, nanoechinate sculpturing of cappa (PRV). (p) Cedrus
sp., EV. (q) Pinus subgenus Pinus sp., EV. (r) Pinus subgenus Strobus sp., EV. (s–t) Tsuga
sp. 1, (s) PV, (t) monosaccus and corpus detail, PRV. (u–v) Tsuga sp. 2, (u) PV, (v)
monosaccus and corpus detail, PRV.
Abbreviations: equatorial view (EV), polar view (PV), distal view (DV), proximal view
(PRV). Scale bars 10 µm (LM, h, t, v), 1 µm (o).
Figure 4. LM and SEM micrographs of Poales, Vitales, Rosales, Fagales, Malpighiales, and
Geraniales.
(a) Typha sp, tetrad, PV. (b–c) Poaceae gen. indet., EV, (c) exine detail, PRV. (d–e)
Monocotyledone indet., (d) PV, (e) PRV. (f–g) Parthenocissus sp., EV. (h) Ulmus vel Zelkova
sp., PV. (I) Fagus sp., EV. (j–k) Quercus sect. Cerris sp., EV, (k) SEM detail, mesocolpium
exine sculpturing. (l–m) Quercus sect. Ilex sp., EV, (m) SEM detail, mesocolpium exine
sculpturing. (n–o) Quercus sect. Quercus sp., PV, (o) SEM detail, apocolpium exine
sculpturing. (p–q) Castanoideae gen. indet. sp., EV, (q) SEM detail, mesocolpium exine
sculpturing. (r) Carya sp., PV. (s) Platycarya sp., PV. (t) Engehardioideae gen. indet., PV. (u)
Alnus sp., PV. (v) Betula sp., PV. (w) Carpinus sp., PV. (x) Corylus sp., PV. (y) Salix sp.,
EV. (z–aa) Geranium sp., (z) PV, (aa) clavae detail.
Abbreviations: equatorial view (EV), polar view (PV), proximal view (PRV). Scale bars 10
µm (LM, e, g), 1 µm (c, k, m, o, q, aa).
Figure 5. LM and SEM micrographs of Sapindales, Malvales, Caryophyllales, Cornales,
Asterales, Dipsacales, and Apiales.
(a–b) Cotinus sp, EV. (c–d) Pistacia sp., (c) PV, (d) exine SEM detail. (e–f) Acer sp. 1, (e)
PV, (f) mesocolpium SEM detail. (g–h) Acer sp. 2, (g) PV, (h) mesocolpium SEM detail. (i–j)
Craigia sp., (i) PV, (j) apocolpium SEM detail. (k) Amaranthaceae gen. indet. sp. 1. (l)
Amaranthaceae gen. indet. sp. 2. (m–n) Caryophyllaceae gen. indet. sp. (o–p) Nyssa sp., (o)
PV, (p) exine sculpturing and aperture SEM detail. (q–r) Fraxinus sp., (q) EV, (r)
mesocolpium SEM detail. (s–t) Olea sp., EV. (u) Cichorioideae gen. indet. sp., PV. (v)
Asteroideae gen indet. sp. 1, PV. (w) Asteroideae gen indet. sp. 2, PV. (x–z) Valeria sp., (x–
y) PV, (z) aperture SEM detail. (aa–bb) Apiaceae gen. indet. sp. 1, EV. (cc–dd) Apiaceae gen.
indet. sp. 2, EV. (ee–ff) Angiosperm pollen fam. et gen. indet. sp., (ee) EV, (ff) mesocolpium
SEM detail.
Abbreviations: equatorial view (EV), polar view (PV). Scale bars 10 µm (LM, b, n, t, y, z, bb,
dd), 1 µm (d, f, h, j, p, r, ff).
Figure 6. Coexistence-Approach diagram showing coexistence intervals for MAT and
CMMT. MAT and CMMT climate ranges of relict taxa a priori excluded from the analysis
are shown on the left side of the diagram.
Blue bars, coldest month mean temperature; red bars, 10–90 percentile climatic range; dark
red extensions, full climatic range.
Figure 7. Köppen signal diagram for the macrofossil and pollen floras of Vegora.
To test and illustrate the stability of the climatic signal, gymnosperms (common alpine
elements) and azonal elements (e.g. riparian or swamp vegetation) were excluded in some
runs.
Supplementary Material
22
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
Supplementary Material Tables S1 - Climatic parameters of NLR
Table S1 (1). Fossil species and climatic parameters of the corresponding NLR (depending on
the fossil-species and their botanical affinities, climate parameters of species, sections,
subgenera, genera, or subfamilies are used as NLR).
Table S1 (2). Climatic parameters of NLR.
Supplementary Material Tables S2 - Köppen-Geiger climate type signatures.
Table S2 (1). Scored Köppen-Geiger signatures of all NLR species of the macrofossil and
pollen flora of Vegora.
Tables S2 (2). Köppen-Geiger signature values and diagram of the macrofossil and pollen
flora of Vegora.
Supplementary Material S3 - Systematic palaeobotany and descriptions of palynomorphs
from the plant fossil bearing strata of Vegora (sample S115992).
Supplementary Material Tables S4 - Palynomorph abundance of sample S115992.
Supplementary Material S5 - Coding of leaf physiognomic characters for morphotypes from
the Vegora lignite mine macroflora. Output PDF files from online CLAMP analysis
(http://clamp.ibcas.ac.cn).
Supplementary Material S6 - Köppen-Geiger categories
23
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
Figure 1.
Figure 1. Fossil localities and lithological map of the Florina–Ptolemais–Servia Basin.
Map redrawn after Steenbrink et al. (1999, 2006), Ognjanova-Rumenova (2005), Ivanov
(2001) and Koufos (2006). Fossil localities: (1) Bitola Basin, Republic of North Macedonia,
PF. (2) Vegora Basin, MF and PF (3) Dytiko, VF. (4) Prosilio, MF. (5) Lava, MF. (6)
Likoudi, MF. (7) Serres Basin. (2–7) Greece. (8) Sandanski Graben, Bulgaria, PF.
Abbreviations: Plant macrofossils (MF), palynoflora (PF), vertebrate fossils (VF).
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
Figure 2.
Figure 2. Lithology and polarity zones of the Vegora section (redrawn after Steenbrink et al.,
2006). Position of fossil bearing strata following Velitzelos and Schneider (1979) and Kvaček
et al. (2002).
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
Figure 3.
Figure 3. Light microscopy (LM) and scanning electron microscopy (SEM) micrographs of
algae, fern and fern allies, and gymnosperm palynomorphs.
(a) Botryococcus sp. cf. B. kurzii. (b) Spirogyra sp. 1/ Ovoidites elongatus. (c) Spirogyra sp.
2/Cycloovoidites cyclus. (d–e) Osmunda sp., (d) EV, (e) PV. (f) Cryptogramma vel
Cheilanthes sp, PV. (g–h) Pteris sp., (g) PV, (h) DV. (i) Davalliaceae vel Polypodiaceae sp./
Verrucatosporites alienus (R.Potonié) P.W.Thomson et Pflug, 1953, EV. (j)
Leavigatosporites haardti, EV. (k–l) Inaperturopollenites hiatus. (m) Abies sp., EV. (n–o)
Cathaya sp., (n) PV, (o) SEM detail, nanoechinate sculpturing of cappa (PRV). (p) Cedrus
sp., EV. (q) Pinus subgenus Pinus sp., EV. (r) Pinus subgenus Strobus sp., EV. (s–t) Tsuga
sp. 1, (s) PV, (t) monosaccus and corpus detail, PRV. (u–v) Tsuga sp. 2, (u) PV, (v)
monosaccus and corpus detail, PRV.
Abbreviations: equatorial view (EV), polar view (PV), distal view (DV), proximal view
(PRV). Scale bars 10 µm (LM, h, t, v), 1 µm (o).
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
Figure 4.
Figure 4. LM and SEM micrographs of Poales, Vitales, Rosales, Fagales, Malpighiales, and
Geraniales.
(a) Typha sp, tetrad, PV. (b–c) Poaceae gen. indet., EV, (c) exine detail, PRV. (d–e)
Monocotyledone indet., (d) PV, (e) PRV. (f–g) Parthenocissus sp., EV. (h) Ulmus vel Zelkova
sp., PV. (I) Fagus sp., EV. (j–k) Quercus sect. Cerris sp., EV, (k) SEM detail, mesocolpium
exine sculpturing. (l–m) Quercus sect. Ilex sp., EV, (m) SEM detail, mesocolpium exine
sculpturing. (n–o) Quercus sect. Quercus sp., PV, (o) SEM detail, apocolpium exine
sculpturing. (p–q) Castanoideae gen. indet. sp., EV, (q) SEM detail, mesocolpium exine
sculpturing. (r) Carya sp., PV. (s) Platycarya sp., PV. (t) Engehardioideae gen. indet., PV. (u)
Alnus sp., PV. (v) Betula sp., PV. (w) Carpinus sp., PV. (x) Corylus sp., PV. (y) Salix sp.,
EV. (z–aa) Geranium sp., (z) PV, (aa) clavae detail.
Abbreviations: equatorial view (EV), polar view (PV), proximal view (PRV). Scale bars 10
µm (LM, e, g), 1 µm (c, k, m, o, q, aa).
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
Figure 5.
Figure 5. LM and SEM micrographs of Sapindales, Malvales, Caryophyllales, Cornales, Asterales, Dipsacales,
and Apiales.
(a–b) Cotinus sp, EV. (c–d) Pistacia sp., (c) PV, (d) exine SEM detail. (e–f) Acer sp. 1, (e) PV, (f) mesocolpium
SEM detail. (g–h) Acer sp. 2, (g) PV, (h) mesocolpium SEM detail. (i–j) Craigia sp., (i) PV, (j) apocolpium
SEM detail. (k) Amaranthaceae gen. indet. sp. 1. (l) Amaranthaceae gen. indet. sp. 2. (m–n) Caryophyllaceae
gen. indet. sp. (o–p) Nyssa sp., (o) PV, (p) exine sculpturing and aperture SEM detail. (q–r) Fraxinus sp., (q) EV,
(r) mesocolpium SEM detail. (s–t) Olea sp., EV. (u) Cichorioideae gen. indet. sp., PV. (v) Asteroideae gen indet.
sp. 1, PV. (w) Asteroideae gen indet. sp. 2, PV. (x–z) Valeria sp., (x–y) PV, (z) aperture SEM detail. (aa–bb)
Apiaceae gen. indet. sp. 1, EV. (cc–dd) Apiaceae gen. indet. sp. 2, EV. (ee–ff) Angiosperm pollen fam. et gen.
indet. sp., (ee) EV, (ff) mesocolpium SEM detail.
Abbreviations: equatorial view (EV), polar view (PV). Scale bars 10 µm (LM, b, n, t, y, z, bb, dd), 1 µm (d, f, h,
j, p, r, ff).
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
Figure 6.
Figure 6. Coexistence-Approach diagram showing coexistence intervals for MAT and
CMMT. MAT and CMMT climate ranges of relict taxa a priori excluded from the analysis
are shown on the left side of the diagram.
Blue bars, coldest month mean temperature; red bars, 10–90 percentile climatic range; dark
red extensions, full climatic range.
bioRxiv preprint doi: https://doi.org/10.1101/848747. this version posted November 25, 2019. The copyright holder for this preprint (which was
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
Figure 7.
Figure 7. Köppen signal diagram for the macrofossil and pollen floras of Vegora.
To test and illustrate the stability of the climatic signal, gymnosperms (common alpine
elements) and azonal elements (e.g. riparian or swamp vegetation) were excluded in some
runs.
bioRxiv
https://doi.org/10.1101/848747
November
25, 2019. The copyright holder for this preprint (which was
Table 1. Plant
taxapreprint
recordeddoi:
from
unit 1 (lignite seam) and unit 2. this
(blueversion
marls) posted
of the Vegora
section.
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
Vegora micro, meso and macro flora
Taxon
Element
Reference
Life form
Ecology
Vegetation units
[BLD subgroups]a
Algae
Botryococcaceae
Botryococcus sp.
P
11
Algae
Aquatic
VU1
Zygnemataceae
Spirogyra spp.
P
11
Algae
Aquatic
VU1
P
L
11
2, 8
Fern
Fern
Swamp, riparian, well-drained lowland forest
Swamp, riparian, well-drained lowland forest
VU3, VU4, VU5
VU3, VU4, VU5
Polypodiaceae
Cryptogramma vel Cheilanthes
P
11
Fern
Swamp, riparian, well-drained lowland forest
VU3, VU4, VU5
Pteridaceae
Pteris sp.
P
11
Fern
Swamp, riparian, well-drained lowland forest
VU3, VU4, VU5
Fam. incerta sedis
Davalliaceae vel Polypodiaceae
P
11
Fern
Swamp, riparian, well-drained lowland forest
VU3, VU4, VU5
P
11
Fern
Swamp, riparian, well-drained lowland forest
VU3, VU4, VU5
Gymnosperms
Ginkgoaceae
Ginkgo adiantoides
L
L
8, 9
Treegym
Riparian forests
VU4
Cupressaceae
Cupressus rhenana
L, R
6, 8
Treegym
Conifer forest lowland, upland, peat-forming
VU7
Sequoia abietina
L
8, 10
Tree gym
Conifer forest lowland, upland, peat-forming
VU7
Cryptomeria anglica
L
10
Treegym
Conifer forest lowland, upland, peat-forming
VU7
Glyptostrobus europaeus
L, R
6, 8
Treegym
Swamp forest
VU3
Taxodium dubium
L, R
6, 8
Tree gym
Swamp forest
VU3
P
11
Tree gym
Indifferent
Fern and fern allies
Osmundaceae
Osmunda sp.
Osmunda parschlugiana
Laevigatosporites haardti
Papillate Cupressaceae
Pinaceae
Abies sp.
P
11
Treegym
Well-drained lowland and upland forests
VU5, VU6, VU7
Cathaya sp.
P
11
Treegym
Conifer forest lowland, upland, peat-forming
VU7
Cedrus sp.
P
11
Treegym
Well-drained lowland and upland forests
VU7
Cedrus vivariensis
R
4, 8
Treegym
Well-drained lowland and upland forest
VU7
Keteleeria hoehnei
Pinus hampeana (diploxylon)
R
6
Tree gym
Conifer forest lowland, upland, peat-forming
VU7
R
6, 8
Tree gym
Well-drained lowland forest
VU5
Pinus salinarum (diploxylon)
R
4, 8
Tree gym
Well-drained lowland forest
VU5
L, R
8
Tree gym
Indifferent
Pinus spp.
Pinus sp. diploxylon type
P
11
Tree gym
Indifferent
Pinus sp. haploxylon type
P, R
8, 11
Tree gym
Indifferent
Pinus vegorae (haplox.)
R
4, 6
Tree gym
Well-drained lowland forest
VU5
Tsuga spp.
P
11
Tree gym
Conifer forest lowland, upland, peat-forming
VU7
Angiosperms
Cabombaceae
Brasenia sp.
R
4
Herb
Aquatic
VU1
Lauraceae
Daphnogene pannonica
L
8
Tree
Well-drained lowland forest
VU5 [BLD wet]
Laurophyllum pseudoprinceps
L
8
Tree
Well-drained lowland forest
VU5 [BLD wet]
Sassafras ferrettianum
Laurophyllum sp.
L
L
7, 8
8
Tree
Tree
Riparian, well-drained lowland forest
Indifferent
VU4, VU5 [BLD wet]
[BLD wet]
Potamogetonaceae
Potamogeton sp.
R
4
Herb
Aquatic
VU1
Arecaceae
Chamaerops humilis fossilis
L
1, 8
Palm
Well-drained lowland forest or scrub
VU0, VU5 [BLD drought]
S
4
Herb
Bogs, wet meadows
VU2
P
11
Herb
Aquatic, bogs, swamp forest, riparian forest
VU1, VU2, VU3, VU4
P
11
Herb
Indifferent
S
3, 4
Herb
Meadows
Zingiberaceae
Spirematospermum wetzleri
Typhaceae
Typha sp.
Poaceae
Poaceae gen. indet.
Cyperaceae
Bolboschoenus vegorae
VU2
Cladium bioRxiv preprint doi: https://doi.org/10.1101/848747
.Herb
this version
posted
November 25, 2019. The copyright holder
S
4
Bogs,
wet meadows
VU2for this preprint (which was
Ceratophyllaceae
Ceratophyllum sp.
not certified by peer review) is the author/funder. It is made available under a CC-BY-NC 4.0 International license.
S
4
Herb
Aquatic
VU1
Platanaceae
L
8
Tree
Riparian, well-drained lowland forest
VU4, VU5 [BLD drought]
Vitaceae
Parthenocissus sp.
Platanus leucophylla
P
11
Liana
Swamp forest, riparian, well-drained lowland forest
VU3, VU4, VU6
Fabaceae
Leguminosites sp.
L
8
Tree
Indifferent
[BLD]
Ulmus plurinervia
P
L
11
8
Tree
Tree
Riparian, well-drained lowland forest
Riparian, well-drained lowland forest
VU0, VU4, VU5 [BLD]
VU4, VU5 [BLD cold]
Zelkova zelkovifolia
L
8
Tree
Mediterranean scrub, riparian, well-drained lowland forest
VU0, VU4, VU5 [BLDdrought]
Ulmaceae
Ulmus vel Zelkova
Fagaceae
Castaneoideae gen indet.
Castanea sp.
Fagus sp.
P
11
Tree
Well-drained lowland forest
VU5 [BLD]
R
P
6, 8
11
Tree
Tree
Well-drained lowland forest
Well-drained lowland and upland forest
VU5 [BLD]
VU5, VU5 [BLD cold]
L, R
6, 7, 8
Tree
Well-drained lowland and upland forest
VU5, VU6 [BLD cold]
P
11
Tree
Well-drained lowland forest
VU5 [BLD]
Quercus cerrisaecarpa
Quercus gigas
R
L
6, 8
7, 8
Tree
Tree
Well-drained lowland forest
Well-drained lowland forest
VU5 [BLD]
VU5 [BLD]
Quercus kubinyii
Quercus sect. Ilex
L
8
Tree
Well-drained lowland forest
VU5 [BLD drought]
P
11
Tree
Mediterranean scrub, well-drained lowland forest
VU0, VU5 [BLD drought]
Quercus drymeja
L
7, 8
Tree
Well-drained lowland and upland forests
VU5, VU6 [BLD drought]
Quercus mediterranea
L
7, 8
Tree
Mediterranean scrub, well-drained lowland forest
VU0, VU5 [BLD drought]
Quercus sosnowskyi
L
7, 8
Tree
Well-drained lowland forest
VU5 [BLD drought]
Quercus sect. Quercus
Quercus pseudocastanea
P
L
11
8
Tree
Tree
Riparian forest, well-drained lowland forest
Well-drained lowland forest
VU4, VU5 [BLD]
VU5 [BLDcold]
Quercus sp.
R
8
Tree
Indifferent
[BLD]
Juglandaceae
Carya sp.
Platycarya sp.
P
P
11
11
Tree
Tree
Riparian, well-drained lowland forest
Riparian, well-drained lowland forest
VU4, VU5 [BLD]
VU4, VU5 [BLD]
L
P
7, 8
11
Tree
Tree
Riparian, well-drained lowland forest
Riparian, well-drained lowland forest
VU4, VU5 [BLD]
VU4, VU5 [BLD]
Fagus gussonii
Quercus sect. Cerris
Pterocarya paradisiaca
Engelhardioideae gen. indet.
Betulaceae
Alnus sp.
P
11
Tree
Swamp, riparian forest, well-drained lowland forest
VU3, VU4, VU5 [BLD cold]
Alnus adscendens
L
8
Tree
Well-drained lowland forest
VU5 [BLD cold]
Alnus cecropiifolia
Alnus cf. kefersteinii
L
8
Tree
Swamp, riparian forest
VU3, VU4 [BLD cold]
R
6, 8
Tree
Indifferent
[BLD cold]
Alnus ducalis
L
7, 8
Tree
Well-drained lowland forest
VU5 [BLD cold]
Alnus gaudinii
L
8
Tree
Well-drained lowland forest
VU5 [BLD cold]
Alnus julianiformis
Betula sp.
L
8
Tree
Riparian, well-drained lowland forest
VU4, VU5 [BLD cold]
P
11
Tree
Riparian, well-drained lowland and upland forest
VU4, VU5, VU6 [BLD cold]
Betula pseudoluminifera
Carpinus sp.
L
8
Tree
Well-drained lowland forest
VU5 [BLD cold]
P
11
Tree
Well-drained lowland forest
VU5 [BLD]
Carpinus betulus fossilis
Carpinus grandis
Carpinus kisseri, group of C. tschonoskii
Corylus sp.
R
L
R
P
8
8
6, 8
11
Tree
Tree
Tree
Shrub
Well-drained
Well-drained
Well-drained
Well-drained
VU5
VU5
VU5
VU5
L
8
Tree
Riparian forest
VU4 [BLD]
L
L
P
8
7, 8
11
Tree
Tree
Tree
Riparian forest
Riparian forest
Swamp, riparian forest
VU4 [BLD]
VU4 [BLD]
VU3, VU4 [BLD]
P
11
Herb
Steppe, meadows, well-drained lowland forest
VU0, VU2, VU5
R
4
Shrub
Swamp
VU3
P
8, 11
Shrub
Well-drained lowland forest
VU5 [BLD drought]
P
11
Tree, shrub (Mediterranean) scrub, well-drained lowland forest
VU0, VU5
Acer aegopodifolium
Acer integrilobum
Acer limburgense (sect. Macrophylla )
L
L
R
7, 8
8
6, 8
Tree
Tree
Tree
Well-drained lowland and upland forests
Well-drained lowland and upland forests
Well-drained lowland and upland forests
VU5, VU6 [BLD]
VU5, VU6 [BLD]
VU5, VU6 [BLD]
Acer pseudomonspessulanum
Acer pyrenaicum (sect. Rubra)
L
8
Tree
(Mediterranean) scrub, well-drained lowland forest
VU0, VU5 [BLD drought]
L
7, 8
Tree
Well-drained lowland forest
VU5 [BLD]
Salicaceae
Populus balsamoides
Populus populina
Populus spp.
Salix sp.
Geraniaceae
Geranium sp.
lowland
lowland
lowland
lowland
forest
forest
forest
forest
[BLD]
[BLD]
[BLD]
[BLD]
Lythraceae
Decodon globosus
Anacardiaceae
Cotinus sp. (=Dicotylophyllum sp. 5)
Pistacia sp.
Sapindaceae
bioRxiv preprint doi: https://doi.org/10.1101/848747
.Tree
this version
posted November
2019.
this[BLD
preprint
(which
was
Acer subcampestre
L
8
Well-drained
lowland and25,
upland
forestThe copyright holder
VU5,for
VU6
]
drought
not certified
by peer review)
is the
author/funder.
It is made available under a CC-BY-NC 4.0 International
license.
Acer tricuspidatum (sect. Rubra
)
L
8
Tree
Swamp, well-drained lowland and upland forest
VU3, VU5, VU6 [BLD cold]
Acer spp.
Malvaceae
Craigia sp.
P, R
8, 11
Tree
Indifferent
[BLD]
P
11
Tree
Well-drained lowland forest
VU5 [BLD]
Craigia bronnii
Droseraceae
R
6, 8
Tree
Well-drained lowland forest
VU5 [BLD]
Aldrovandia praevesiculosa
Caryophyllaceae
Caryophyllaceae gen. indet. sp.
Amaranthaceae
S
4
Herb
Aquatic
VU1
P
11
Herb
Steppe, meadows, well-drained lowland and upland forest
VU0, VU2, VU5, VU6
Amaranthaceae/Chenopodiaceae gen. indet. sp.
Nyssaceae
Nyssa sp.
P
11
Herb
Steppe, meadows
VU0, VU2
P
11
Tree
Swamp, well-drained lowland and upland forest
VU3, VU5, VU6 [BLD wet]
R, P
P
8, 11
11
Tree
Tree
Riparian forest
Mediterranean scrub, well-drained lowland forest
VU4 [BLD]
VU0, VU5
P
11
Herb
Steppe, meadows, well-drained lowland and upland forest
VU0, VU2, VU5, VU6, VU7
P
11
Herb
Steppe, meadows, well-drained lowland and upland forest
VU0, VU2, VU5, VU6, VU7
P
11
Herb
Steppe, meadow, riparian
VU0, VU2, VU4
L
7, 8
Liana
Riparian, well-drained lowland forest
VU4, VU5
P
11
Herb
Steppe, meadows, well-drained lowland and upland forest
VU0, VU2, VU5, VU6, VU7
Monocotyledone indet.
Dicotylophyllum sp. 1–4, 6 (oder spp.)
L
L
8
8
Herb
Swamp, riparian, lake margin
VU2, VU3, VU4
Monocotyledone indet.
Pollen indet.
P
P
11
11
Oleacaee
Fraxinus sp.
Olea sp.
Asteraceae
Asteraceae gen. indet. spp.
Cichorioideae gen. indet. sp.
Caprifoliaceae
Succisa sp.
Araliaceae
Hedera multinervis
Apiaceae
Apiaceae gen. indet. spp.
Incerta sedis
a
BLD (Broadleaf deciduous forest biome of Woodward et al., 2004) was divided in BLD drought, deciduous trees and shrubs that are drought resistant, and BLD cold,
deciduous trees and shrubs that are cold tolerant; in addition we use BLD wet for trees and shrubs that typically occur in humid warm temperate regions.
Vegetation Unit (VU) 0: Steppe-like meadows with shrubs and/or small trees scattered or in groups; Mediterranean scrub. VU 1: Aquatic. VU 2: Bogs, wet meadows.
VU 3: Swamp forest. VU 4: Riparian forest. VU 5: Well-drained lowland forest -a ‘‘hot’’(Lauraceae, Chamaerops, Engelhardioideae, Olea );
-b ‘‘temperate’’ (Castanea , Carpinus , Tilia) including levee forests. VU 6: Well-drained upland forest (-a Quercus drymeja-mediterranea ; -b Fagus-Cathaya ).
VU 7: Well-drained (lowland and) upland conifer forest including hammocks and raised bogs within peat-forming vegetation.
L = leaves; P = palynomorph; R = reproductive structures
1) Velitzelos & Schneider, 1979; 2) Velitzelos & Petrescu, 1981; 3) Velitzelos et al., 1983; 4) Velitzelos & Gregor, 1985; 5) Mai & Velitzelos, 1992; 6) Mai & Velitzelos, 1997;
7) Velitzelos & Kvaček, 1999; 8) Kvaček et al., 2002; 9) Denk & Velitzelos, 2002; 10) Velitzelos & Denk, 2002; 11) this study
Table 2. Estimated climate parameters for the pre-evaporitic Messinian of Vegora from two CLAMP calibration datasets and from CA.
Climate parameter
CLAMP
CLAMP
CA modified
CA modified
Physg3arcAZ PhysgAsia1
10-90%iles
MAT (°C)
10–13.5
8.7–11.5
8.6–21.2
9.9–18.4
CMMT (°C)
1–5
-2.7–2.3
≥ 1.2
WMMT (°C)
19.2–22.8
19–22.6
GROWSEAS (months)
6–8
5.5–7
MMGSP (mm)
110–160
100–160
Three_WET (mm)
500–780
400–750
Three_DRY (mm)
180–260
80–220
3_WET/3_DRY
<4
< 5.5
MAT = mean annual temperature, CMMT = coldest month mean temperature,
WMMT = warmest month mean temperature, GROWSEAS = duration of growing season,
MMGSP = mean month growing season precipitation, Three_WET = precipitation of
three consecutive wettest months, Three_DRY = precipitation of three consecutive
driest months.
Table 3. Vegetation types recognised for the pre-evaporitic Messinian of the Florina–Ptolemais–Servia Basin.
a
Vegetation type
Swamp forest
Main (and accessory) taxon/taxa
Taxodium , Glyptostrobus
Biome
NLD
Vegetation unit(s)
VU3
Swamp forest
Riparian forest
BLD
BLD
VU3
VU4
BLD
VU5b
Well-drained forest
Alnus , (Sassafras )
Pterocarya , Zelkova , Ulmus ,
(Sassafras )
Quercus kubinyi , Q. pseudocastanea ,
(Carpinus , Tilia etc.)
Fagus , (Quercus pseudocastanea )
BLD
VU5b
Well-drained forest
Fagus , Abies , Cedrus , Cathaya
MIXED
VU6b
Well-drained laurophyllous forest
Quercus drymeja ,
(Q. sosnowsky )
Quercus mediterranea ,
Chamaerops , Olea
Poaceae
BLE
VU6a
Well-drained forest
Well-drained sclerophyllous
forests/shrublands
[?]c Grassland-steppe forest
b
Modern (Neogene) analogue
Taxodium swamp forests SE USA;
(Taxodium/Glyptostrobus swamp forests
widespread in N Hemisphere Neogene)
Alnus swamp forest
Riparian and alluvial forest of Georgia
and Iran
Lowland oak-hornbeam forests;
("Quercetum mixtum")
Lowland beech forests of N Turkey,
Georgia, N Iran; ("Fagetum gussonii")
Montane Fagus-Abies forest,
montane Fagus-Cedrus-Pinus forest;
Abant Gölü; Erbaa-Çatalan
Quercus dilatata association (with Taxus , Pinus ,
Acer etc.)
Mediterranean sclerophyllous
forest/shrublands
Forest-steppe of SE Europa to Afghanistan
BLE/
VU0
SHRUBLAND
GRASSLAND/ VU0
SHRUBLAND
1) Mai, 1995; 2) Dolezych & Schneider, 2007; 3) Denk et al., 2001; 4) Akhani et al., 2010; 5) Maharramova, 2015; 6) Kozlowski et al., 2018;
7) Kvaček et al., 2002; 8) Mayer & Aksoy, 1986; 9) Akkemik, 2003; 10) van Zeist & Bottema, 1991; 11) Freitag, 1971; 12) Erdős et al., 2018
a
Biome classification follows the phsiognomic approach of Woodward et al., 2004. bVegetation units as in Table 1.
c
[?] expresses the uncertainty around a possible extra-regional signal in the Vegora pollen record. According to Erdös et al. (2018) steppe forest with Stipa and
other grasses and different species of Quercus (forest-steppes of the type 'Region A - SE Europe') is characterized by MAP of 420-600 mm;
this would be much drier than the inferred MAP for the FPS.
References
1, 2
3, 4
1, 3, 4, 5, 6
4, 7
3, 4, 7
8, 9, 10
11
10
8, 10, 11, 12