American Journal of Botany
Eocene Araucaria Sect. Eutacta from Patagonia and floral turnover during the initial isolation of South America --Manuscript Draft--
Manuscript Number: AJB-D-19-00445R2 Full Title: Eocene Araucaria Sect. Eutacta from Patagonia and floral turnover during the initial isolation of South America Short Title: ROSSETTO-HARRIS ET AL. — PATAGONIAN ARAUCARIA SECT. EUTACTA FOSSILS Article Type: Research Article Section/Category: Paleobotany Corresponding Author: Gabriella Rossetto-Harris, M.S. Pennsylvania State University University Park, PA UNITED STATES Corresponding Author E-Mail: [email protected] First Author: Gabriella Rossetto-Harris, M.S. Order of Authors: Gabriella Rossetto-Harris, M.S. Peter Wilf Ignacio H. Escapa Ana Andruchow-Colombo Abstract: PREMISE OF THE STUDY— Eocene floras of Patagonia document biotic response to the final separation of Gondwana. The conifer genus Araucaria, distributed worldwide during the Mesozoic, has a disjunct extant distribution between South America and Australasia. Fossils assigned to Australasian Araucaria Sect. Eutacta usually are represented by isolated organs, making diagnosis difficult. Araucaria pichileufensis Berry, from the middle Eocene Río Pichileufú (RP) site in Argentine Patagonia, was originally placed in Sect. Eutacta and later reported from the early Eocene Laguna del Hunco (LH) locality. However, the relationship of A. pichileufensis to Sect. Eutacta and the conspecificity of the Araucaria material among these Patagonian floras have not been tested using modern methods. METHODS— We review the type material of A. pichileufensis alongside large (n = 192) new fossil collections of Araucaria from LH and RP, including multi-organ preservation of leafy branches, ovuliferous complexes, and pollen cones. We use a total evidence phylogenetic analysis to analyze relationships of the fossils to Sect. Eutacta. KEY RESULTS— We describe Araucaria huncoensis sp. nov. from LH and improve the whole-plant concept for Araucaria pichileufensis from RP. The two species respectively resolve in the crown and stem of Sect. Eutacta. CONCLUSIONS— Our results confirm the presence and indicate the survival of Sect. Eutacta in South America during early Antarctic separation. The exceptionally complete fossils significantly predate several molecular age estimates for crown Eutacta. The differentiation of two Araucaria species demonstrates conifer turnover during climate change and initial South American isolation from the early to middle Eocene. Keywords: Araucariaceae; biogeography; conifers; early Eocene climatic optimum; Gondwana; Laguna del Hunco; Patagonia; rainforest; Río Pichileufú; total evidence phylogeny
Funding Information: National Science Foundation Dr. Peter Wilf (DEB-1556666) National Science Foundation Dr. Peter Wilf (EAR-1925755) National Science Foundation Dr. Peter Wilf (DEB-0919071) National Science Foundation Dr. Peter Wilf (DEB-0345750)
Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation Botanical Society of America Mrs. Gabriella Rossetto-Harris (Graduate Student Research Grant) Geological Society of America Mrs. Gabriella Rossetto-Harris (Graduate Student Research Grant) Graduate School, Pennsylvania State Mrs. Gabriella Rossetto-Harris University (Penn State Geosciences Charles E. Knopf, Sr., Memorial Scholarship; and a Penn State Geosciences Paul D. Krynine Scholarship)
Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation Manuscript Click here to access/download;Manuscript;MS Rossetto Harris Revision 2_2.docx
1
1 Eocene Araucaria Sect. Eutacta from Patagonia and floral turnover during the
2 initial isolation of South America1
3
4 Gabriella Rossetto-Harris2,4, Peter Wilf2, Ignacio H. Escapa3, and Ana Andruchow-Colombo3
5
6 1Manuscript received ______2019; revision accepted ______.
7 2Department of Geosciences, Pennsylvania State University, University Park, Pennsylvania
8 16802 USA; and
9 3Museo Paleontológico Egidio Feruglio, Consejo Nacional de Investigaciones Científicas y
10 Técnicas, Trelew 9100, Chubut, Argentina
11 4Author for correspondence (e-mail: [email protected])
12
13 ROSSETTO-HARRIS ET AL. — PATAGONIAN ARAUCARIA SECT. EUTACTA FOSSILS
14
15 PREMISE OF THE STUDY — Eocene floras of Patagonia document biotic response to the
16 final separation of Gondwana. The conifer genus Araucaria, distributed worldwide during the
17 Mesozoic, has a disjunct extant distribution between South America and Australasia. Fossils
18 assigned to Australasian Araucaria Sect. Eutacta usually are represented by isolated organs,
19 making diagnosis difficult. Araucaria pichileufensis Berry, from the middle Eocene Río
20 Pichileufú (RP) site in Argentine Patagonia, was originally placed in Sect. Eutacta and later
2
21 reported from the early Eocene Laguna del Hunco (LH) locality. However, the relationship of A.
22 pichileufensis to Sect. Eutacta and the conspecificity of the Araucaria material among these
23 Patagonian floras have not been tested using modern methods.
24 METHODS — We review the type material of A. pichileufensis alongside large (n = 192) new
25 fossil collections of Araucaria from LH and RP, including multi-organ preservation of leafy
26 branches, ovuliferous complexes, and pollen cones. We use a total evidence phylogenetic
27 analysis to analyze relationships of the fossils to Sect. Eutacta.
28 KEY RESULTS — We describe Araucaria huncoensis sp. nov. from LH and improve the
29 whole-plant concept for Araucaria pichileufensis from RP. The two species respectively resolve
30 in the crown and stem of Sect. Eutacta.
31 CONCLUSIONS — Our results confirm the presence and indicate the survival of Sect. Eutacta
32 in South America during early Antarctic separation. T exceptionallyhe complete fossils
33 significantly predate several molecular age estimates for crown Eutacta. The differentiation of
34 two Araucaria species demonstrates conifer turnover during climate change and initial South
35 American isolation from the early to middle Eocene.
36 KEY WORDS —Araucariaceae; biogeography; conifers; early Eocene climatic optimum;
37 Gondwana; Laguna del Hunco; Patagonia; rainforest; Río Pichileufú; total evidence phylogeny
38
39 The final breakup of Gondwana had a profound influence on conifer evolution and biogeography
40 (McLoughlin, 2001; Leslie et al., 2012; Kooyman et al., 2014). Conifer taxa previously
41 described from the Eocene fossil caldera-lake deposits of Laguna del Hunco (~52.2 Ma), Chubut,
3
42 Argentina, and Río Pichileufú (~47.7 Ma), Río Negro, Argentina (Figure 1; Berry, 1938; Wilf et
43 al. 2009, 2014, 2017a; Wilf, 2012), illustrate a complex biogeographic history of Old World
44 (mostly Australasian/SE Asian) and New World survival patterns following the early Cenozoic
45 isolation of Antarctica and onset of global cooling (Zachos et al., 2001; Lauretano et al., 2018).
46 Laguna del Hunco preserves a highly diverse late Gondwanan flora (Wilf et al., 2003,
47 2013; Appendix S1; see Supplemental Data with this article) from the early Eocene climatic
48 optimum (EECO), the longest period of sustained warmth in the Cenozoic (~53-49 Ma; Zachos
49 et al., 2001; Lauretano et al., 2018). In that greenhouse world, the westernmost edge of a frost-
50 free, trans-Antarctic, mesothermal rainforest biome reached into southern South America via a
51 close connection between Patagonia and Antarctica (e.g., Hill 1995; Wilf et al. 2013; Kooyman
52 et al., 2014). The Río Pichileufú flora (Berry, 1938; Wilf et al., 2005) represents a period of
53 initial climatic cooling and drying following the EECO, as known from regional and global
54 records (Pearson et al., 2007; Hollis et al., 2012; Bijl et al., 2013; Dunn et al., 2015), from a time
55 when the earliest separation of South America and Antarctica was underway (e.g., Lawver et al.,
56 2011). Comparison of the well-studied elements of the Laguna del Hunco and Río Pichileufú
57 floras (Appendix S1), yields evidence of turnover in the angiosperms, cycads, and ferns,
58 suggesting some extinction and a shift in the paleoenvironment (Wilf et al., 2013, 2019).
59 However, until now there has been no detection of a significant change in species composition of
60 the prevalent conifers between the two localities, and abundant Araucaria fossils at both sites are
61 considered to have affinities with A. Sect. Eutacta (e.g., Wilf et al. 2017a) but have not been
62 formally investigated since the 1940s (Berry 1938; Florin 1940).
63 The family Araucariaceae consists of the sister clades Araucaria Juss. and the agathioid
64 clade, comprised of the sister genera Agathis Salisb. and Wollemia W.G. Jones, K.D. Hill & J.M.
4
65 Allen (Gilmore and Hill, 1997; Stefanovic et al., 1998; Kunzmann, 2007; Liu et al., 2009; Leslie
66 et al., 2012; Escapa and Catalano, 2013; Escapa et al., 2018). Araucaria includes 20 extant
67 species within four sections that show a disjunct distribution: South American Section Araucaria
68 and Australasian Sects. Bunya, Eutacta, and Intermedia (Endlicher, 1847; Wilde and Eames,
69 1952; Farjon, 2010; Mill et al., 2017). Section Eutacta Endlicher (New Caledonia, New Guinea,
70 Australia) accounts for most of the extant diversity within Araucaria, encompassing 16 species
71 (Farjon, 2010; Mill et al., 2017).
72 Araucaria species are large, monoecious (rarely dioecious: Sect. Araucaria) evergreen
73 trees that are often recognized by their characteristic candelabra-shaped, pyramidal, or columnar
74 crowns (Farjon, 2010). Their leaves are spirally arranged and can be differentiated into juvenile
75 and adult forms (Sects. Intermedia and Eutacta ; Fig. 2A-C, F-H; Wilde and Eames, 1952; de
76 Laubenfels, 1972, 1988; Farjon, 2010). Leaves are multi-veined (Sects. Araucaria, Bunya, and
77 Intermedia) or single-veined (Sect. Eutacta), with margins entire or minutely denticulate, and
78 amphistomatic (Wilde and Eames, 1952; de Laubenfels, 1972, 1988; Stockey and Ko, 1986;
79 Farjon, 2010). Stomata in Araucaria lack Florin rings, and features such as stomatal orientation,
80 size, and the number of subsidiary cells are diagnostic for the sections and species (Florin, 1931;
81 Stockey and Taylor, 1978; Stockey and Ko, 1986).
82 Araucaria pollen cones are subtended by a cluster of sterile, triangular, pointed basal
83 bracts that emerge directly at the cone base and extend along the cone body (Fig. 2E; de
84 Laubenfels, 1972; Farjon, 2010). Eutacta is the only section with terminal pollen cones (Fig. 2C;
85 de Laubenfels, 1988), which are otherwise axillary in the genus. Ovuliferous complexes (OCs) in
86 Araucaria have a single inverted seed embedded in their central tissues. The OC often extends
87 laterally as thin to woody wings and has a narrow spine that projects above the convex thickened
5
88 apical margin, termed the apophysis (Fig. 2I, J, K). A ligule is distal to the seed and reaches the
89 apophysis (Fig. 2K); the ligule is interpreted as the distal free tip of the ovuliferous scale, which
90 is almost completely fused to the bract (de Laubenfels, 1988). Section Eutacta has small,
91 samara-like ovuliferous complexes, with broad lateral bract margins that are thin and papery
92 (Fig. 2J; de Laubenfels, 1988).
93 Araucaria has a famously rich fossil history worldwide, in contrast to its modern,
94 fragmented Southern Hemisphere range, and the genus was highly diverse during the Mesozoic
95 (Stockey, 1982, 1994; Hill and Brodribb, 1999; Kershaw and Wagstaff, 2001; Kunzmann, 2007;
96 Panti et al., 2012). Mesozoic fossils with varying degrees of completeness have been suggested
97 to represent Araucaria section Eutacta (i.e., Seward, 1903; Kendall, 1949 ; Archangelsky, 1966;
98 Bose and Maheshwari, 1973; Harris, 1979; Cantrill, 1992; Pole, 1995; Cantrill and Falcon-Lang,
99 2001; Axsmith et al., 2008; van der Ham et al., 2010), but there is no consensus on their
100 relationships (Stockey, 1994; Leslie et al., 2012 ). The majority of Cenozoic macrofossil records
101 of Araucaria Sect. Eutacta are based on single organs (see Table 1). Although their affinities to
102 the section are more reliable than the Mesozoic fossils due to better preservation, the lack of
103 multiple organs still makes it difficult to be certain of many fossils’ phylogenetic positions.
104 Araucaria ligniticii, from the Oligocene of Victoria, Australia (Cookson and Duigan, 1951), is
105 one of the few Cenozoic Eutacta species that is based on multiple organs, including ovuliferous
106 complexes, leaves, and a fragment of a pollen cone.
107 Molecular estimates of the stem divergence of Section Eutacta range from ~57 to ~145
108 Ma (Leslie et al., 2012, 2018), depending on the calibrations and methods used, contrasting with
109 some fossil evidence suggesting a minimum age of ca. 190 Ma (Axsmith et al., 2008). Molecular
110 crown ages of Sect. Eutacta are usually post-Gondwanan, ca. 20 Ma (Biffin et al., 2010; Leslie et
6
111 al., 2012, 2018), or between ca. 51 and ca. 9 Ma (Kranitz et al., 2014). The young molecular age
112 estimates of Sect. Eutacta would suggest evolution of the clade entirely after Gondwanan
113 separation and a wholly Old World biogeographic history (Crisp and Cook, 2011; Kranitz et al.,
114 2014), whereas fossil ages provide direct evidence that many southern conifer clades were
115 present in West Gondwana prior to the separation of landmasses and had vastly greater ranges
116 than they do now (as summarized by Wilf and Escapa, 2015). The post-Gondwanan molecular
117 ages of Araucaria Sect. Eutacta can be directly tested using well-preserved Gondwanan fossils
118 with affinities to Eutacta.
119 Of interest here, the Río Pichileufú site from the early middle Eocene of Río Negro,
120 Argentina is the type locality for Araucaria pichileufensis Berry, described from one ovuliferous
121 complex and five leafy-branch fossils (Berry, 1938). Berry (1938) and Florin (1940) placed A.
122 pichileufensis within Sect. Eutacta. Modern excavations at Río Pichileufú (Wilf et al., 2005,
123 2017a) have significantly increased the sampling of A. pichileufensis (fossil material shown in
124 Figs. 3–10), including the discoveries reported here of previously undescribed pollen cones, one
125 of these attached to a leafy branch; in situ fossil pollen; and, rarely, in situ leaf cuticle, greatly
126 increasing the number of characters that can be used to diagnose the fossils. Notably, fossil
127 Araucaria from Laguna del Hunco, which has not been formally studied until now, has also been
128 thought to represent A. pichileufensis in the absence of any previously observed differences from
129 the Río Pichileufú material (Wilf et al., 2005, 2017a). The abundant Laguna del Hunco
130 Araucaria material also includes preservation of leafy branches, OCs, and pollen cones (Wilf et
131 al., 2014, 2017a).
132 This paper will (1) investigate whether modern taxonomic treatment and new fossil
133 material of Araucaria pichileufensis supports its placement in A. Section Eutacta as suggested
7
134 long ago (Berry, 1938; Florin 1940); (2) test the conspecificity of the Araucaria fossils from
135 Laguna del Hunco and Río Pichileufú; and (3) evaluate the systematic placement of the fossils
136 using a total evidence phylogenetic analysis. We will also explore the implications of these
137 results for the biogeographic history of Section Eutacta and for floral turnover in Patagonia in
138 the context of the middle Eocene onset of South American isolation and global cooling.
139
140 MATERIALS AND METHODS
141
142 The Araucaria pichileufensis fossils presented here were found at Río Pichileufú, Río Negro,
143 Argentina, and the fossils of a new species of Araucaria came from Laguna del Hunco, Chubut,
144 Argentina. Both the Río Pichileufú and Laguna del Hunco localities are fossiliferous caldera-lake
145 deposits of the Eocene Huitrera Formation (Petersen, 1946; Aragón and Romero, 1984; Aragón
146 and Mazzoni, 1997; Iannelli et al., 2017). Berry (1925, 1938) initially interpreted the two sites to
147 be Miocene in age and similar in species composition. Berry did not report Araucaria from
148 Laguna del Hunco; the first mention was by Petersen (1946).
149 Over the past two decades, significant field collections have resulted in numerous
150 descriptions and revisions of fossil plants, insects, ichnotaxa, and vertebrates from Río Pichileufú
151 and Laguna del Hunco, summarized previously ( e.g., Wilf et al., 2009, 2013, 2014; see also
152 Appendix S1 ). Paleomagnetic stratigraphy at Laguna del Hunco revealed two paleomagnetic
153 reversals within the 170 meters of section, and three tuffs that interbed the 31 fossil quarries have
154 been radiometrically dated (Wilf et al., 2003, 2005; Wilf 2012) . At Laguna del Hunco, abundant
155 Araucaria fossils are found above and below volcanic ash 2211A, which yielded sanidine
156 crystals with an analyzed 40Ar-39Ar age of 52.22 ± 0.22 Ma (Wilf et al., 2005, 2017b) . The other
8
157 two ashes produced similar ages (Wilf et al. 2003), and thus the working age for the Laguna del
158 Hunco flora is 52.2 Ma (early Eocene, Ypresian; Wilf et al., 2017b). Araucaria is found in
159 quarries LH2, LH4, LH6, LH13, LH15, LH20, LH22, LH23, LH25, LH27, and LH28–LH31
160 (Wilf et al., 2003, 2005; Gandolfo et al., 2011). New quarries from the 2019 field season include
161 LH30, which is at the same level as LH6; and LH29 and LH31, which are in the lower section,
162 near LH13. Araucaria is most abundant at the LH13 and LH27 quarries (see Systematics).
163 At Río Pichileufú, three tuffs containing sanidines produced concordant 40Ar-39Ar ages,
164 resulting in a combined analytical age of 47.74 ± 0.05 Ma (Wilf et al., 2005, Wilf 2012); these
165 tuffs are located just above the fossiliferous strata, which include abundant Araucaria fossils.
166 The age of the Río Pichileufú flora is thus most likely earliest middle Eocene (Ypresian/Lutetian
167 boundary is 47.8 Ma; Gradstein et al., 2012 as updated at www.stratigraphy.org). The new
168 Araucaria pichileufensis fossils reported here were found at quarries RP1–3 (Wilf et al., 2005),
169 and at a new quarry from the 2017 field season, RP4, which is at the same stratigraphic level as
170 RP3.
171 Laguna del Hunco and Río Pichileufú are both highly diverse, angiosperm-dominated
172 floras with more than 100 species found at each locality ( e.g., Wilf et al., 2005). Elevated species
173 richness from these floras is probably linked to abundant, aseasonal rainfall and frost-free
174 winters at the time (Wilf et al., 2009). Appendix S1 summarizes the taxa from Laguna del Hunco
175 and Río Pichileufú that have been studied systematically in recent years or can otherwise be
176 considered reliable, and their presence, absence, and relative abundance between the two
177 localities. Gymnosperms show less apparent turnover than angiosperms between Laguna del
178 Hunco and Río Pichileufú (Appendix S1). Huncocladus laubenfelsii, an extinct podocarp related
179 to Phyllocladus known from a single specimen (Andruchow-Colombo et al., 2019), is the only
9
180 Laguna del Hunco conifer species that is not known from Río Pichileufú. Although the conifer
181 diversity is almost consistent across the sites, most conifers are less abundant at Río Pichileufú
182 than Laguna del Hunco (Appendix S1).
183 Fossil repositories, preparation, and imaging — The historic Río Pichileufú type and
184 cohort collections of Araucaria pichileufensis (Berry 1938) are housed in the Paleobotanical
185 Division of the Smithsonian Institution, National Museum of Natural History (USNM). The
186 remainder of the Patagonian fossils reported here were collected during several expeditions
187 based out of the Museo Paleontológico Egidio Feruglio (MEF, Trelew, Chubut, Argentina;
188 repository acronym MPEF-Pb) from 1999-2019 and the Universidad Nacional del Comahue (San
189 Carlos de Bariloche, Río Negro, Argentina) in 2017. Fossils collected on these fieldtrips are
190 housed at the Museo Paleontológico de Bariloche ( San Carlos de Bariloche, Río Negro,
191 Argentina; repository acronym BAR) for Río Pichileufú material and at MEF for Laguna del
192 Hunco. The total collection of fossil Araucaria analyzed here includes 136 specimens from
193 Laguna del Hunco and 56 specimens from Río Pichileufú.
194 Type and cohort Araucaria pichileufensis collections (26 specimens) were loaned from
195 USNM to be examined at the Paleobotany Laboratory, Pennsylvania State University (PSU).
196 Material from BAR was loaned to MEF, and all except the historical (USNM) collections were
197 prepared, imaged, and analyzed at MEF. Preparation of fossils to remove extraneous matrix was
198 done using airscribes. Macroimages were taken using a Nikon D90 camera with 60 mm macro
199 lens (Nikon, Melville, New York, USA), a polarizing filter, and low angle, unidirectional light.
200 Fossils were also examined using reflected light microscopy (at MEF, Nikon Eclipse 50i
201 compound microscope; at PSU, Nikon Eclipse LV100 compound microscope and Nikon SMZ-
202 1500 binocular scope), and images of fine details were taken using an attached camera (at MEF,
10
203 Nikon DSFi3 with an L4 control unit; at PSU, Nikon DS-Ri1 and NIS Elements BR software).
204 For observation of fossil cuticle and pollen, a Nikon C-SHG1 epifluorescence illumination unit
205 was used on the MEF Eclipse microscope, and an X-Cite 120 epifluorescence illumination unit
206 (EXFO Electro-Optical Engineering, Quebec City, Quebec, Canada) was used on the PSU
207 Eclipse microscope and camera setup. Images at multiple depths of field were stacked using
208 Adobe Photoshop CC (version 20.0.1; Adobe Inc., San Jose, California, USA). The Photoshop
209 tools Auto-Align, Auto-Blend, and Photomerge were used to synthesize stacked and laterally
210 stitched images.
211 We attempted to remove coal and expose underlying cuticle from the Laguna del Hunco
212 foliar specimens using bleach (NaClO 5%) on in situ and separated cuticle fragments. The
213 cuticle treatments were unsuccessful, but it was still possible to observe unprepared cuticle in
214 situ using epifluorescence. Pollen was observed in situ in a single Araucaria pichileufensis cone
215 specimen from Río Pichileufú. We did not attempt to separate pollen from the cone for analysis
216 under SEM because there are only three pollen cone specimens from the locality, and destructive
217 analysis would be required. O bservation of the in situ pollen under epifluorescence was
218 sufficient to observe the characters needed for analysis.
219 Extant material — Living Araucaria Sect. Eutacta collections were examined at the
220 Montgomery Botanical Center, Coral Gables, Florida, and vouchers are deposited at the
221 Pennsylvania State University Herbarium (PAC; Appendix S2). In addition to the living
222 collections, extant Araucaria species were examined and photographed using the same
223 macrophotography methods as above, from herbarium collections at the Herbarium of the Arnold
224 Arboretum, Harvard University Herbaria (A, Cambridge, Massachusetts, USA) and The
225 Huntington Library, Art Collections, and Botanical Gardens (HNT, San Marino, California,
11
226 USA), where the late D. J. de Laubenfels donated part of his research collection in 2010. High-
227 resolution images of herbarium sheets were also accessed online from Muséum National
228 d’Histoire Naturelle, Paris (P;
229 http://science.mnhn.fr/institution/mnhn/collection/p/item/search/form). Some of the resulting
230 images of specimens from the PAC and A collections are published here with permission (Figure
231 2).
232 Analysis of specimens— All images of fossil and herbarium specimens were annotated
233 and organized using a common set of keywords in Adobe Bridge CC (version 9.0.1.216; San
234 Jose, California, USA) to expedite comparison of a unified set of morphological characters in a
235 large sample (after Wilf et al. 2017a). Full-resolution images of fossil specimens are deposited
236 open-access at FigShare (see Data Availability). Measurements of specimens are given in Table
237 1. For the leafy branches, the leaf length and width are given for the abaxial view of the
238 imbricate/adpressed foliage forms (see Systematics). For the length of curved pollen cone
239 specimens, measurements were estimated using a straight line from base to apex and found to be
240 similar to lengths estimated by segmenting along the cone length. Measurements of fossils were
241 made virtually on scaled photographs using Adobe Photoshop and cross-checked on specimens
242 with calipers. Araucaria species taxonomy follows Farjon (2010), de Laubenfels (1972, 1988),
243 and Mill et al. (2017).
244 Box plots to analyze the ovuliferous complex measurements were made using the
245 webtool BoxPlotR ( Spitzer et al., 2014). The raw data were tested with graphical and analytical
246 methods and fit the assumptions needed to use a t test to measure the significance of category
247 differences: independent observations, normality, and homogeneity of variables (for length:width
248 ratios of OCs, log transformation was needed to normalize the data).
12
249 Phylogenetic analysis— The two Eocene Araucaria species in this study were included
250 in a previously published character matrix, slightly revised here, to analyze their phylogenetic
251 positions (Escapa and Catalano, 2013; Escapa et al., 2018; Appendix S3) . The total-evidence
252 approach simultaneously combined (1) a morphological matrix for an ingroup composed of 32
253 living and 12 fossil species of Araucariaceae, including the two fossils analyzed here for the first
254 time, and an outgroup composed of 10 extant species from the conifer families Cupressaceae,
255 Pinaceae, and Podocarpaceae; and (2) a molecular dataset for the aforementioned 32 living
256 ingroup species and 306 outgroup species from five extant conifer families (Podocarpaceae,
257 Sciadopityaceae, Cupressaceae, Taxaceae, and Pinaceae). The morphological matrix comprises
258 53 discrete characters and 10 continuous characters (standardized in TNT prior to the analysis,
259 using the command nstates stand). The molecular dataset assembled by Escapa and Catalano
260 (2013) and Escapa et al. (2018) includes 23 molecular markers, including plastid, nuclear, and
261 mitochondrial gene fragments. The complete Escapa et al. (2018) dataset can be downloaded
262 online as their appendix S1, and the TNT file used in the present analyses, including the scores
263 for the two new fossil species analyzed here, is available as Appendix S3 ( the TNT file can be
264 opened in any text reader application) and online as a project linked to this article on
265 MorphoBank (see Data Availability; O’Leary and Kaufman, 2011). Besides the addition of
266 scores for the two fossil Araucaria species studied here, scores for extant Araucaria species were
267 changed for character 37 (pollen cone morphology) and character 40 (basal bracts on pollen
268 cone); these species were previously scored as unknown (?) in the original matrix (Escapa and
269 Catalano 2013). Here, those characters are scored based on data from Farjon (2010) and
270 observations of herbarium specimens.
13
271 The fossils were first removed from the combined matrix to test the methodology on only
272 the extant taxa, after which fossils were included. Combined phylogenetic analyses of the dataset
273 were performed using equally weighted parsimony in TNT v.1.5 (Goloboff et al., 2003, 2008;
274 Goloboff and Catalano, 2016). The analysis was run using Sectorial Search, Drift, and Tree
275 Fusing, set to stop after finding the minimum length 30 times independently. Then, tree
276 bisection-reconnection (TBR) was used with the trees previously found, and the strict consensus
277 tree was calculated. Group support values were calculated with absolute bootstrap frequencies
278 (standard sampling with replacement) for 500 replicates.
279 The molecular-only support values were nearly identical to previous results for major
280 clades, but as shown in many prior studies, the support values were much lower when fossils
281 were included (i.e., Wilkinson, 2003; Escapa and Pol, 2011). The addition of fossils to the
282 analyses introduced uncoded characters, including the missing genetic data and many
283 morphological characters that are not preserved in fossils, all of which lowered the overall
284 support of the tree with all three group-support methods. Within the problematic New
285 Caledonian Eutacta clade, low support values are expected even without the addition of fossils,
286 due to the known genetic and morphologic similarities within the clade (e.g., Setoguchi et al.,
287 1998; Stefenon et al., 2006; Gaudeul et al., 2012, 2014; Ruhsam et al., 2015). Despite the low
288 support values, including the fossils in the analysis resulted in the same topology for Araucaria
289 and Araucariaceae that was found for the molecular-only phylogeny.
290
291 SYSTEMATICS
292
14
293 Family— Araucariaceae J. B. Henkel & W. Hochstetter, Synopsis der Nadelhölzer: xvii (1865).
294
295 Genus— Araucaria de Jussieu, Genera Plantarum: 413 (1789).
296
297 Species— Araucaria pichileufensis E.W. Berry, Geological Society of America Special Paper
298 12: 59 (1938).
299
300 Lectotype, here designated— USNM 40383e (Fig. 7A; Berry 1938: plate 11, figure 1), an
301 ovuliferous complex from the historic locality of Río Pichileufú, Huitrera Formation, earliest
302 middle Eocene, Río Negro Province, Argentina, exact collection location unknown.
303
304 Syntypes— USNM 40383a (Fig. 3G; Berry 1938, pl. 11, fig.5), USNM 40383b (Berry 1938: pl.
305 11, fig. 2), USNM 40383d (Fig. 3H; Berry 1938, pl. 11, fig. 6), USNM 40383f (Berry 1938, pl.
306 11, fig. 4), USNM 545203–545207, 545209–545222 (not figured by Berry; Fig. 3A, USNM
307 545222; Fig. 3B, USNM 545210; see Remarks). The syntype USNM 40383c (Berry 1938, pl. 11,
308 fig. 3) is here excluded from the species.
309
310 Other material here referred— Río Pichileufú, Huitrera Formation, middle Eocene, Río Negro
311 Province, Argentina. Leafy branch segments with subulate spreading foliage: BAR 4237, 4357
312 (quarry RP1); 5576 (RP4); 5581, 5586 (RP3); 5587 (RP4); 5592, 5596, 5608, 5612, 5614, 5627
313 (RP2). Leafy branch segments with lanceolate adpressed foliage: BAR 4225 (RP1); 5347
15
314 (unknown locality); 5574 (RP3); 5575, 5577, 5579, 5588, 5589 (RP4), 5592, 5602 (RP2); 5609
315 (unknown locality). Ovuliferous complexes: BAR 4352 (RP1); 5407, 5427 (unknown locality);
316 5569, 5580, 5584 (RP4); 5591 (RP1); 5593–5595, 5598, 5599, 5603, 5605, 5607, 5610, 5613
317 (RP2); 5616 (RP4); 5620, 5622–5626, 5628, 5629 (RP2). Pollen cones: BAR 5582 (RP4); 5618,
318 5619 (RP2).
319
320 Emended description— Leafy branches (n = 23; Fig. 3A–J) are preserved as segments 2.5–25.3
321 cm long and 4.7–14.5 mm wide including foliage, ca. 0.6–1.9 mm wide without foliage. Leafy
322 branch segments are straight or slightly curved. The branch apex ends in a cluster of reduced,
323 incompletely formed leaves (Figs. 3C, 3H, 5F).
324 Leaves are helically arranged and sessile. Individual leaves along branches are preserved
325 either in lateral profile view (Fig. 3A–E) or abaxial view (Fig. 3G–J). Both views can occur on a
326 single branch segment (Fig. 3E, F). The profile of the leaf in lateral view shows a combination of
327 the abaxial and adaxial surfaces, which together approximate the height of the leaf. The abaxial
328 view shows an approximation of the true width of the leaf; the probable maximum width toward
329 the leaf base is obscured due to imbrication, as is the leaf length. Leaves are differentiated into
330 longer, keeled, subulate spreading and shorter, adpressed lanceolate types and can vary
331 continuously, or somewhat abruptly, in size and type along the branch segment (Fig. 3C, D, F).
332 One leafy branch specimen (Fig. 3F) appears to have foliage of both types, with some subulate
333 spreading leaves and some lanceolate adpressed leaves in attachment along a single shoot. Leaf
334 insertion angle, measured on leaves preserved in lateral view, varies from ca. 19–48°.
16
335 Subulate spreading foliage type (Fig. 3A–E): Leaves are divergent and are inserted at an
336 average angle of ca. 33° (varies between 22–48°). Leafy branches with this foliage type are 8–14
337 mm wide, including foliage. Leaf shape is subulate (awl-shaped), slender, wider at base and
338 tapering to a point. Seen in lateral view, leaves are abaxially keeled, varying from 3.3–10.2 mm
339 in length and 1.2–2.2 mm in height, and somewhat linear (Figs. 3C–E, 5A) to slightly incurved
340 (Fig. 3A, B). The leaf apex is acute and pointed. Leaves of the spreading type are usually only
341 preserved in lateral view. Rhomboidal leaf bases along the branch are present where leaves have
342 broken off or detached (Fig 3E) and measure ca. 0.9–3.6 mm long and 1.8–2.6 mm wide.
343 Lanceolate adpressed foliage type (Figs. 3F–J, 5B, 5F): Leaves are adpressed and densely
344 packed on leafy branches that are 4.7–8.3 mm wide, including the foliage. Leaf insertion angle is
345 ca. 31° (varies between 19–32°). Leaf shape is lanceolate when viewed completely exposed on
346 the abaxial face (Figs. 3F, 3G, 3I, 5B), and in lateral view, the leaves are recurved toward the
347 central axis. In abaxial view, leaves are ca. 5.0 mm long (3.9–6.7 mm) and ca. 2.9 mm wide
348 (2.2–4 mm). The apex is acute and incurved. Leaf bases where leaves have broken off are ca. 2.3
349 mm wide (1.9–2.9 mm) and 2.4 mm long (2.1–2.7 mm).
350 Cuticle remains were observed on two specimens of the spreading foliage type as
351 coalified compressions (Fig. 6A, E, F). Stomata are arranged in parallel rows and mostly oriented
352 parallel to the long axis of the leaf, with some oblique or perpendicularly oriented (Fig. 6E, F).
353 Leaves are amphistomatic, based on preservation of stomata on a few adaxial leaf surfaces, in
354 addition to more easily observed abaxial surfaces (Fig. 6A). Stomatal apparatus, including
355 subsidiary cells, are ca. 70 µm long and 60 µm wide. At least four subsidiary cells per stoma are
356 observed, with both polar and lateral subsidiary cells present, but the ability to differentiate cell
17
357 wall locations between subsidiaries is limited by preservation. Guard cells are ca. 25 µm long by
358 15 µm wide.
359 Ovuliferous complexes ( n = 25; Figs. 7A, 7C, 7E, 8A, 8C, 8E –H) are samara-like, with a
360 central seed fused into the central body that is surrounded on either side by broad lateral wings
361 that may be incompletely preserved (Figs. 7E, 8A, 8G). The OC is broadly flabellate (fan-
362 shaped), with an average length of 24.6 mm (20.5–26.8 mm), and a maximum width of ca. 28.9
363 mm (23.4–33.4 mm) including the wing-like lateral extensions; maximum width occurs at ca.
364 50% (38–63%) of the total length. The wings are bluntly rounded at the distal lateral margins.
365 The width of the base of the OC averages 9.4 mm (8.1–11.2 mm). The OC has a thickened
366 central apical region, or apophysis, which projects as a convex triangle and ends in a ca. 2 mm
367 long narrow tip or spine (Fig. 7A, B). The average maximum width at the apophysis is 15.4 mm
368 (13.3–17.2 mm), positioned at about 73% (59–80%) of the scale length. A ligule is centered
369 directly above the seed apex (Fig. 8A, C, E). The ligule is long and triangular, ca. 4.3–4.4 mm in
370 length and 4.2–6.7 mm wide at its widest point. The apex of the ligule is located ca. 2 mm from
371 the distal margin of the apophysis, where the spinose apical projection begins.
372 Seeds are inverted, embedded in the ovuliferous complex tissue, and located in the center
373 of the bract-scale complex. Seeds are obovate, narrowing to a point at the micropylar end, ca.
374 15.3 mm long and 6.4 mm wide at maximum width, located at ca. 73% of the seed length (from
375 the micropylar to the chalazal end). The micropylar end of the seed is positioned ca. 2.3 mm
376 (1.3–3.2 mm) from the base of the scale. Seeds are often striated (Fig. 8A) and have amber casts
377 within inferred resin duct locations (Fig. 8F, H).
378 Pollen cones (n = 3; Figs. 9A, 9B, 10A–F) are terminal on leafy branches (attachment
379 preserved in one specimen: Fig. 9A), solitary, cylindrical, and often slightly curved, ending in a
18
380 bluntly rounded apex. The single complete specimen (Fig. 9B) is 55 mm long and 7.4 mm wide;
381 widths of the other specimens are 7.7 mm (Fig. 9A) and 8.3 mm (Fig. 10A). Basal bracts are
382 longer and more triangular than regular leaves, clustered and beginning immediately at the base
383 of the cone body, and extending distally to overlap the base of the cone body (Figs. 9B, 10A,
384 10F). The basal bracts are triangular, 6.4–8.6 mm long and 1.2–1.3 mm wide, and the apices are
385 acute and pointed. Cuticle observed on a basal bract of a pollen cone shows stomata oriented
386 mostly parallel to the long axis of the bract (Fig. 10D), like the corresponding foliage. On the
387 single preserved face of a cone, ca. 5–8 overlapping bracts are visible (Fig. 10F), suggesting that
388 more than ten total were originally present around the full cone base. Microsporophylls are
389 helically arranged and imbricate. The microsporophyll laminae are peltate and rhombic with an
390 acute, pointed apex, measuring ca. 1 mm by 1 mm (Fig. 10E). Microsporophyll stalks (Fig. 10B,
391 E) are thin and reflexed basally. Discernible pollen sacs are not preserved. Estimated number of
392 microsporophylls per cone is ca. 800 (ca. 380 counted on a single face of the complete cone
393 shown in Fig. 9B).
394 Pollen was found in situ in one pollen cone specimen (Fig. 10B, C). The pollen was
395 observed in clumps along the microsporophyll stalks and inside the cone body under
396 epifluorescence (Fig. 10B). The pollen grains are non-saccate, inaperturate, and spherical to sub-
397 spherical, measuring about 60 by 55 µm (Fig. 10C). Because the pollen was observed in situ, it
398 was not clear if the exine is sculptured or smooth.
399
400 Species— Araucaria huncoensis Rossetto-Harris, sp. nov.
401
19
402 Etymology— The specific epithet references the type locality, Laguna del Hunco. “Hunco” is a
403 variant of “junco,” for the rushes (Juncus sp.) that are locally present around the intermittent
404 pond that carries the name (i.e., “Lake of Rushes”).
405
406 Holotype here designated— MPEF-Pb 10556 (Fig. 8B), ovuliferous complex from quarry LH27,
407 Laguna del Hunco, Tufolitas Laguna del Hunco, Huitrera Formation, early Eocene, northwest
408 Chubut Province, Argentina.
409
410 Paratypes— Laguna del Hunco, Tufolitas Laguna del Hunco, Huitrera Formation, early Eocene,
411 northwest Chubut Province, Argentina. Leafy branch segments with acicular falcate foliage:
412 MPEF-Pb 10587–10590, 10592–10594, 10596–10599 (quarry LH13); 10600, 10604 (LH15);
413 10607, 10608 (LH27); 10609 (LH25); 10630 (LH13); 10633 (LH27); 10640 (LH13). Leafy
414 branch segments with ovate imbricate foliage: MPEF-Pb 10580 (LH6); 10581 (LH13); 10582
415 (LH25); 10583–10586 (LH27); 10591 (LH6); 10595 (LH13); 10601 (float specimen); 10602
416 (LH6); 10603 (LH22); 10605 (LH13); 10606 (LH25), 10610–10613 (LH27); 10626 (LH31),
417 10629 (LH29). Ovuliferous complexes: MPEF-Pb 10510 (LH6); 10511, 10512 (LH2); 10513,
418 10514 (LH13); 10515 (LH6); 10516 (LH2); 10517–10520 (LH13); 10521 (LH22); 10522
419 (LH13); 10523 (LH6); 10524–10529 (LH13); 10530 (LH20); 10531 (float specimen); 10532–
420 10534 (LH13); 10535 (LH25); 10536, 10537 (LH27); 10538–10540 (LH13); 10541 (LH6);
421 10542 (LH27); 10543, 10544 (LH13); 10545 (LH27); 10546 (LH20); 10547 (LH13); 10548
422 (LH15); 10549 (LH13); 10550 (LH15); 10551 (LH27); 10552 (LH25); 10553 (LH27); 10554
423 (LH13); 10555 (LH6); 10556, 10557 (LH27); 10558 (LH6); 10559 (LH2); 10560 (LH13); 10561
20
424 (LH22); 10562 (LH2); 10563 (LH27); 10564 (LH2); 10565, 10566 (LH13); 10567–10574
425 (LH27); 10575–10579, 10621–10623 (AL1); 10624 (LH31); 10625 (LH29); 10627, 10628
426 (LH29); 10631 (LH13); 10632, 10634 (LH27); 10635, 10636 (LH13); 10637 (LH29); 10638
427 (LH27); 10639 (LH29); 10640 (LH13); 10641 (LH27); 10642 (LH29); 10643 (LH13); 10644
428 (LH29); 10645 (LH13); 10646, 10647 (LH13); 10648, 10649 (LH28); 10650 (LH13); 10651
429 (LH29); 10652, 10653 (LH13). Pollen cones: MPEF-Pb 10614 (LH27); 10615 (LH6); 10616
430 (LH27); 10617 (LH13); 10618, 10619 (LH27).
431
432 Diagnosis— Ovuliferous complexes are narrowly flabellate-cuneate, with narrow, angular lateral
433 extensions and ratios of maximum length to width close to one. Ovuliferous complex apophysis
434 is rounded, and the ligule is a short, broad triangle. Leafy branches are dimorphic with acicular
435 falcate and ovate imbricate foliage types, stomata perpendicularly oriented to the long axis of the
436 leaf.
437
438 Description— Leafy branches (n = 36; Fig. 4A–J) are long and slender, with preserved segments
439 up to 24.5 cm long, 5.0–12.5 mm wide with foliage, ca. 0.9–2.4 mm wide without foliage.
440 Leaves are helically arranged and sessile; preserved in lateral (Fig. 4A–E, G) and abaxial views
441 (Fig. 4F, H–J); and differentiated into acicular falcate and ovate imbricate types.
442 Acicular falcate foliage type (Figs. 4A–D, 5C, 5I): Leafy branches are 6.0–12.5 mm
443 wide. Leaves are falcate, acicular incurved, 4.3–8.2 mm long and 1.2–2.3 mm high with an acute
444 apex. The leaf insertion angle is ca. 29° (18–37°). Leaf bases where the leaves have broken off of
445 the branch are ca. 2.5–2.7 mm long and 1.9–2.4 mm wide.
21
446 Ovate imbricate foliage type (Fig. 4E–J): Leafy branches are 5–11 mm wide. Leaves are
447 imbricate and tightly adpressed, with an average insertion angle of 29° (ranges between 12–50°).
448 Leaves are broadly ovate (Fig. 5D) and apices are acute (i.e. Fig. 4I). Viewed laterally, leaves are
449 strongly incurved apically, 2.3–5.7 mm long and 1.1–3.0 mm high. In abaxial view, leaves are
450 ca. 4.2 mm long (3.4–6.5 mm) and ca. 3.0 mm wide (2.5–4.8 mm). Depending on the amount of
451 imbrication of leaves, leaf lengths in lateral or abaxial views can be underestimated. Leaf bases
452 remaining on the branch where leaves have broken off are on average 2.5 mm long by 2.6 mm
453 wide (ranging from 2.2–2.7 mm long and 2.4–3.4 mm wide; Figs. 4F, 4J, 5H).
454 Cuticle remains (Fig. 6B–D, G, H) were observed on seven specimens of both foliage
455 types. Stomata are arranged in parallel rows and oriented mostly perpendicular to the
456 longitudinal axis of the leaf, with some parallel or oblique (Fig. 6G, H). Leaves are
457 amphistomatic (Fig. 6B–D). Stomatal apparatus, including subsidiary cells, are ca. 80 µm long
458 and 65 µm wide. The number of subsidiary cells is not distinguishable due to coalification.
459 Guard cells are ca. 50 µm long by 30 µm wide.
460 Ovuliferous complexes ( n = 64; Figs. 7B, 7D, 7F, 8B, 8D, 8I–K ) are samara-like,
461 narrowly flabellate to cuneate, average length 18.1 mm (13.3–24.5 mm), maximum width ca.
462 18.9 mm (13.9–33.9 mm) including narrow, symmetrical lateral extensions that are angled at the
463 lateral distal margins. The position of maximum width is ca. 62% (46–73%) of the total length.
464 Some stomata were observed on the wings of OCs (Fig. 8I, J), oriented in rows and mostly
465 perpendicular to the long axis of the OC like the corresponding foliage. The guard cells range
466 from ca. 60–65 µm in length and are ca. 25 µm wide. Width of the base of the OC averages 6.5
467 mm (3.3–9.9 mm). The OC has a thickened apophysis that projects in a convex rounded profile
468 and ends in a ca. 1.7 mm long narrow tip or spine. The average maximum width at the apophysis
22
469 is 11.7 mm, positioned at ca. 79% (66–89%) of the OC length. The ligule is present ca. 0.7–1.3
470 mm distal to the seed, showing where the scale terminates in a free tip (Fig. 8B, D). The ligule is
471 a broad, low triangle that is 0.6–1.9 mm in height and 3.3–5.3 mm wide. The ligule apex is ca.
472 0.3–07 mm from the distal margin of the apophysis, where the spinose apical projection begins.
473 Seeds are inverted and embedded in the center of the ovuliferous complex. Seeds are
474 obovate, ca. 12.3 mm long and 5.3 mm wide, and the point of maximum width is ca. 76% of the
475 seed length (from the micropylar to the chalazal end). The micropylar end of the seed is located
476 ca. 2 mm (1.2–4.4 mm) from the base of the OC. Seeds are often striated (Figs. 7F, 8B, 8D, 8K).
477 Pollen cones (n = 7; Figs. 9C, 9D, 10G, 10H) are cylindrical and often strongly curved,
478 all found isolated. A complete specimen (Fig. 9D) is 86 mm long and 8.8 mm wide, but
479 incomplete specimens can reach >105 mm in length and range from ca. 7.5–10 mm wide (Figs.
480 9C, 10H). Leaf-like bracts are clustered at the base of the cone and overlap the base of the cone
481 body for the entire length of the bract (Fig. 10G, H). Basal bracts are triangular, 9.0–12.5 mm
482 long and 0.3–0.5 mm wide, and the apex is acute and pointed (Fig. 10G). On a single preserved
483 cone face ca. 4–5 bracts are visible, suggesting that more than eight were present in total on the
484 living cone (Fig. 10H). Microsporophylls are helically arranged, imbricate, and peltate (Fig. 9C),
485 with a rhombic external face and acute pointed apex, the face ca. 2 mm long by 1.7 mm wide.
486 Stalks are thin and reflexed basally (Fig. 9C). Pollen sacs and pollen are not preserved. Estimated
487 number of microsporophylls per cone is ca. 900 (ca. 450 counted on a single face of the one
488 complete cone: Fig. 9D).
489
23
490 Remarks—We chose an ovuliferous complex as the lectotype of Araucaria pichileufensis
491 because it is the organ that bears the greatest number of characters that distinguish the species,
492 including the overall broadly flabellate shape, the wide, rounded lateral extensions, and the
493 triangular apophysis. We excluded one syntype, USNM 40383c (Berry 1938: pl. 11, fig. 3), from
494 A. pichileufensis because we re-identified that specimen as Dacrycarpus puertae due to its small,
495 bilaterally flattened leaves deployed in a single plane; juvenile leaves of Araucaria Sect. Eutacta
496 are bifacially flattened, acicular, tetragonal in cross-section, and never twisted into a horizontal
497 plane (de Laubenfels, 1988). The listed, originally nonfigured specimens are also considered
498 here as syntypes because they were part of the original gathering that was studied by Berry
499 (1938), as shown by their identification tags in his handwriting, and reside as a cohort collection
500 to the figured types at USNM. Because there was no original diagnosis (Berry 1938), there is
501 nothing to emend formally , but the diagnostic characters of A. pichileufensis are its subulate,
502 spreading and lanceolate, adpressed dimorphic leaf forms with parallel orientation of stomata
503 and its broadly flabellate ovuliferous complexes with rounded wings, triangular apophyses, and
504 long, triangular ligules.
505 Assignment of the two fossil species to Araucariaceae— The foliage of extant
506 Araucariaceae is spirally arranged or opposite-decussate and can be needle-like, scaly, or broad
507 (de Laubenfels, 1988). The foliage of both fossil species (Figs. 3–5) is spirally arranged and
508 needle-like to scaly, conforming to the range of morphologies for the family. The ovuliferous
509 complexes from this family, as in the fossils described here (Figs. 7, 8), have a single inverted
510 seed and a bract that is almost completely fused with the scale. The presence of isolated fossil
511 OCs that vary continuously in size suggests that the scales of both fossil species were once
512 spirally arranged around an ovuliferous cone (Fig. 2I) and dispersed at maturity, as is
24
513 characteristic of the extant family. Both A. pichileufensis and A. huncoensis are also assigned to
514 Araucariaceae based on cylindrical pollen cones (Figs. 9, 10), in the former species found
515 attached to the foliage (Fig. 9A), that have spirally arranged, imbricate, peltate microsporophylls
516 and bear subtending basal bracts (de Laubenfels, 1988; Gilmore and Hill, 1997). The surface
517 texture of in-situ pollen grains on an A. pichileufensis pollen cone (Fig. 10C) could not be
518 observed; however, based on their general features, including their rounded, non-saccate form
519 and inaperturate structure, the grains are consistent with Araucariaceae.
520 Assignment to Araucaria— The leafy branches of the two fossil species can be assigned
521 to Araucaria on the combined basis of their crowded, spiral (helical) phyllotaxy (Figs. 3, 4),
522 scale to needle-like morphology with broad, sessile attachment to the branch (Fig. 5A–D),
523 amphistomy, absence of Florin rings, and stomatal arrangement in discontinuous rows (Fig. 6A–
524 D; de Laubenfels, 1953). Leaves of extant and fossil Agathis species, including Agathis
525 zamunerae from the same localities, are, in contrast, in well-separated pairs, (sub)opposite along
526 the branch, broad and oval or elliptic with a narrow false petiole, and hypostomatic (de
527 Laubenfels, 1953). Agathis leafy branches also differ from Araucaria because they have distinct
528 terminal buds with overlapping scales (as seen in Agathis zamunerae), whereas Araucaria have
529 only a cluster of reduced leaves at the branch apex (Fig. 5F, I; de Laubenfels, 1988). The leaves
530 of the fossils described here also differ significantly from Wollemia, which has trimorphic,
531 usually four-ranked, broad leaves with obtuse to rounded apices that are opposite to subopposite
532 and twisted (Jones et al., 1995; Farjon, 2010).
533 The fossil ovuliferous complexes are typical of Araucaria, with a single inverted seed
534 embedded in central tissues (Figs. 7, 8) and a distal ligule (Figs. 2J, K,8A-E; Wilde and Eames,
535 1952; de Laubenfels, 1972, 1988; Stockey, 1994; Farjon, 2010). The presence of broad,
25
536 symmetrical OCs that are inferred to be thinned laterally into wing-like extensions, with a
537 narrow, spinose projection from the apex above a thickened apical margin (Fig. 2J), is
538 characteristic of Araucaria (Figs. 7, 8; de Laubenfels, 1988). In contrast, Agathis and Wollemia
539 have OC s that lack a ligule and are thus interpreted as having their bract and scale completely
540 fused (Jones et al., 1995). Seeds of Wollemia and Agathis, including Agathis zamunerae, are
541 dehiscent and winged; their OCs are completely lignified, lack a projecting apical spine, and
542 have characteristic basal embayments (Farjon et al., 2010; Wilf et al., 2014). All these features
543 are unlike the fossils presented here.
544 The pollen cones of both fossil species (Figs. 9, 10) are typical of those from extant
545 Araucaria (Fig. 2E) based on the morphology of the basal bracts, which occur in dense clusters
546 overlapping the cone body and are triangular to lanceolate with sharp apices, broad at the base,
547 and flattened (de Laubenfels, 1972, 1988). The basal bracts on pollen cones of Agathis
548 (including A. zamunerae) occur below or barely overlap the cone body and are distinctly short,
549 rounded, and ovate, occurring in only three pairs (Farjon, 2010; Wilf et al., 2014).
550 Microsporophyll apices of Agathis are rounded; however, the fossils described here, like living
551 Araucaria, have pointed, acute apices (Fig. 10E; Wilf et al., 2014). Wollemia differs from the
552 fossils described here in having basal bracts that are broadly triangular to semicircular (Jones et
553 al., 1995) and do not subtend the fallen cones (as seen in the fossils), as well as rounded and
554 clavate microsporophyll apices (Chambers et al., 1998).
555 Overall, the leafy branches, ovuliferous complexes, and pollen cones of both fossil
556 species have multiple diagnostic characters of Araucaria, and there is no extinct genus that is
557 similar to the fossils, so there is no need to assign them to an extinct genus. For example, A.
558 pichileufensis and A. huncoensis both differ from the extinct genus Araucarioides, which has
26
559 multi-veined leaves that are flattened and strap-like, with obliquely oriented stomata (Bigwood
560 and Hill, 1985; Pole, 2008).
561 Assignment to Section Eutacta— The leafy branches of both fossil species are similar in
562 morphology to those of extant Sect. Eutacta and exhibit foliar dimorphism compatible with the
563 differentiated juvenile and adult foliage of the section (Fig. 2A–C, F–H). The adult leaves in
564 extant Sect. Eutacta have small, scale -like, imbricate, single -veined leaves (Figs. 2C, 2H; Wilde
565 and Eames, 1952; de Laubenfels, 1972; Farjon, 2010) that are consistent with the adpressed
566 leaves present in both fossil species (Figs. 3F–J, 4E–F, 5B, 5 D). All other sections have large,
567 flattened, multiveined, spreading leaves (Farjon 2010). The juvenile leaves of extant Sect.
568 Eutacta are typically awl-shaped or needle-like (Figs. 2A, 2E ; Wilde and Eames, 1952; de
569 Laubenfels, 1972; Farjon, 2010), which is consistent with the acicular and spreading foliage
570 types present in the fossil species (Figs. 3A–E, 4A–D, 5C, 5I). Among extant Araucaria, only
571 Sects. Eutacta and Intermedia have species with foliage that is considered differentiated into
572 juvenile and adult states (de Laubenfels, 1988; Farjon, 2010), although seasonal dimorphism,
573 involving differences in size and shape for broad, multiveined leaves, has been observed for
574 Sects. Bunya and Intermedia (Cantrill and Falcon-Lang, 2001; Andruchow-Colombo et al.,
575 2018). Typically, the stomata in mature leaves of Sect. Eutacta are oriented obliquely or
576 perpendicularly to the long axis of the leaf (Stockey and Ko, 1986). The fossil stomata seen on
577 Araucaria huncoensis are perpendicular (Fig. 6G), so they are consistent with Eutacta. However,
578 the small number of stomata observed in fossils of A. pichileufensis were mostly oriented parallel
579 to the long leaf axis (Fig. 6E), a character state that is shared with the other three extant sections
580 of Araucaria (Stockey and Ko, 1986) and thus could represent a plesiomorphy (see Discussion).
27
581 The fossil ovuliferous complexes of both species show a single seed that appears to be
582 embedded in the surrounding tissues (Figs. 7, 8). The central scale portion, including the seed
583 and apophysis region, is often darker in color (more coalified) compared to the lateral bract
584 wings, suggesting that the wings were markedly thinner than the central body as is characteristic
585 of Sect. Eutacta. Sometimes, only the central body is preserved, and the OC appears to be
586 lacking wings entirely (Fig. 8G) or is narrower than the majority of complete specimens (Fig.
587 8A). The incomplete preservation of the wings and variation in wing size and shape is probably
588 taphonomic or due to breakage before preservation, as observed in extant Sect. Eutacta (see Fig.
589 2J). The OC fossils are very different from extant Araucaria bidwillii (Sect. Bunya), which are
590 large, completely woody, and have round, dehiscent seeds, and from extant Sect. Araucaria,
591 which has OCs that are large and nut-like, with completely reduced lateral wings. The fossils
592 differ from extant Sect. Intermedia in size and shape, being much smaller than the broad, semi-
593 circular, fan-shaped, membranous OCs of Araucaria hunsteinii, and in their apices, which are
594 blunt and rhombic in A. hunsteinii (de Laubenfels, 1988).
595 The terminal attachment of pollen cones seen in Araucaria pichileufensis (Fig. 9A) is,
596 within extant Araucaria, exclusively a character of Section Eutacta (Fig. 2C). All three other
597 sections have axillary pollen cones (de Laubenfels, 1988; Farjon, 2010).
598 Whole plant hypotheses — For the Araucaria fossils at Río Pichileufú and Laguna del
599 Hunco, we hypothesize that the associated leafy branch, ovuliferous complex, and pollen cone
600 organs originated from a single respective species at each locality. All the Araucaria plant
601 organs described here are found together in the same fossil horizons at their respective sites, and
602 they are clearly distinguishable from the associated Agathis zamunerae fossils in each organ
603 category (as described here and informally by Wilf et al., 2014). Variation within each of the
28
604 organs is within range for a single species, and the respective pairs of foliage types co-occur at
605 the same stratigraphic level at Río Pichileufú and throughout the stratigraphic section at Laguna
606 del Hunco; thus, there is no evidence of more than one Araucaria species at either locality. At
607 Río Pichileufú, one specimen showed leaves of both types along a single branch (Fig. 3F). The
608 pollen cone organically attached to a leafy branch at Río Pichileufú (Fig. 9A) confirms that the
609 dispersed pollen cones with triangular basal bracts found in the same fossil horizons as the
610 isolated leafy branches at Río Pichileufú are also part of the Araucaria pichileufensis plant. In
611 addition, the parallel stomatal orientation on a basal bract of a pollen cone from Río Pichileufú
612 (Fig. 10D) is equivalent to that of A. pichileufensis leaves (Fig. 6E). Similarly, cuticle seen on
613 ovuliferous complexes from Laguna del Hunco had perpendicular stomatal orientations (Fig. 8J)
614 that were also equivalent to those of leaves at that locality (Fig. 6G). Although not observed in
615 situ in Laguna del Hunco pollen cones, dispersed fossil pollen newly reported from the site,
616 assigned to Araucariacites australis, probably corresponds to the A. huncoensis macrofossils
617 (Barreda et al., in press).
618 Differentiation of two fossil species— Informal references to “Araucaria pichileufensis”
619 at Laguna del Hunco have appeared in several papers (e.g., Petersen 1946; Wilf et al., 2005,
620 2014). However, A. pichileufensis and A. huncoensis are here well separated based on
621 morphological differences in both vegetative and reproductive organs (Table 1; Figs. 5–12),
622 justifying the new species from Laguna del Hunco. In Araucaria pichileufensis, the dimorphic
623 leaf forms are (1) subulate and spreading and (2) lanceolate and adpressed, whereas in Araucaria
624 huncoensis, the dimorphic forms are (1) acicular and falcate, and (2) ovate and imbricate. The
625 ovate and imbricate leaves of A. huncoensis are wider and slightly shorter than the lanceolate and
626 adpressed leaves of A. pichileufensis (Figs. 3, 4, and 5). Stomata are oriented differently between
29
627 the two species; A. pichileufensis has parallel orientation of stomata relative to the leaf axis (n =
628 3; Fig. 6E), and A. huncoensis has perpendicular stomata (n = 7; Fig. 6G).
629 There are several significant differences in the dimensions of the ovuliferous complexes
630 among the two species (Fig. 11). The broad, rounded lateral wings of A. pichileufensis contribute
631 to the large maximum width of the OCs, which is, on average, about five millimeters greater than
632 the total length. In contrast, the lateral wings of A. huncoensis are much narrower, with a
633 maximum width that is on average almost identical to the total length, making the overall outline
634 more cuneate (Fig. 11A–C). The A. huncoensis OCs have a length to width ratio that is close to
635 one, whereas A. pichileufensis OCs are generally wider than they are long (Fig. 11C). The A.
636 pichileufensis OCs have maximum widths of ca. 28.7 mm, positioned at ca. 48% of the OC
637 length, whereas the maximum width of the lateral wings of A. huncoensis is ca. 18.9 mm,
638 positioned at ca. 63% of the total length (Fig. 11 B, D). At the apophysis, A. pichileufensis is
639 more triangular, but A. huncoensis is more rounded. Araucaria pichileufensis has broader base
640 widths than A. huncoensis (ca. 9.5 mm in A. pichileufensis vs. ca. 6.5 mm in A. huncoensis ), but
641 the ratio of the maximum width to the base width for both species is about the same. Ligule size
642 also differs between the two fossil species. The OCs of A. pichileufensis have longer, triangular
643 ligules (Fig. 8A, C), but those of A. huncoensis have short, broad triangular ligules (Fig. 8B, D).
644 Based on the single complete specimen for each species, pollen cones of A. huncoensis
645 are longer than those of A. pichileufensis. The A. huncoensis pollen cones are also on average
646 wider and have slightly larger external microsporophyll faces.
647 Comparison with other fossils— Some Mesozoic fossils based solely on ovuliferous
648 complexes could belong to Section Eutacta (see Introduction). The two Mesozoic fossils with
649 possible affinity to Eutacta that include multi-organ preservation are, unsurprisingly, different
30
650 from the Eocene fossils presented here. Brachyphyllum mamillare Lindley & Hutton and
651 associated Araucarites phillipsii Carruthers from the Jurassic of Yorkshire (Kendall, 1949;
652 Harris, 1979) have short leaves (only reaching lengths of 1.5–4.0 mm), small OCs (15 mm by 13
653 mm), and small pollen cones (only 12 mm long and 6 mm wide). From the Jurassic-Cretaceous
654 boundary interval in Argentina (Baldoni, 1979), Brachyphyllum feistmantelii (Halle) Sahni leafy
655 branches are associated with small OCs (10 mm long and 8 mm wide) of Araucarites chilensis
656 Baldoni and pollen cones associated with Brachyphyllum feistmantelii (12–13 mm long and 6–9
657 mm wide).
658 Among Paleogene fossils assigned to Sect. Eutacta (Table 1), Araucaria cf. A.
659 pichileufensis was reported from Pampa de Jones, Neuquén, Argentina, an early Eocene site that
660 is slightly older than Laguna del Hunco and also from the Huitrera Formation (Wilf et al., 2010:
661 fig. 3b). The single illustrated OC is much more similar to Araucaria huncoensis than to A.
662 pichileufensis and probably belongs to A. huncoensis. Although incompletely preserved, the
663 Pampa de Jones specimen has a small maximum width (at least 15 mm) and cuneate shape like
664 A. huncoensis, with narrow lateral extensions on either side of the seed. The ligule is similar to
665 those observed at Laguna del Hunco but with a shorter length.
666 Araucaria fildesensis, from the early-middle Eocene Fossil Hill Formation of King
667 George Island, Antarctica, is based on a single OC (Shi et al., 2018). Compared to the OCs
668 described here, the A. fildesensis OC is kite-shaped, rather than flabellate. The Antarctic fossil
669 appears to lack any thin lateral extensions on either side of the seed, making it narrower in
670 overall width. Although the margins of its OC appear fairly straight, it is likely that the wings
671 were either broken off before preservation or incompletely preserved, based on the variability
672 seen in the living material of Sect. Eutacta (i.e. Fig. 2J). Associated “juvenile” and “mature”
31
673 Araucaria foliage (Zhou and Li 1994; Shi et al. 2018), some of which has been assigned to Sect.
674 Eutacta (Shi et al. 2018), differs significantly from A. pichileufensis and A. huncoensis. The awl-
675 shaped “juvenile” Antarctic leaves are widely inserted at 45–70°, only slightly incurving toward
676 the branch, whereas the leaves of A. huncoensis are more incurving, inserted at angles of ca. 29°.
677 Leaves of “juvenile” Antarctic Araucaria sp. are also much more widely spaced on the branches
678 than the Patagonian leaves, allowing for a clear view of the branch that is rarely seen on any
679 specimen from Río Pichileufú or Laguna del Hunco. The “juvenile” Antarctic leaves are much
680 longer and narrower compared with the fossils described here and are uniformly acicular rather
681 than lanceolate. The imbricate, keeled and lanceolate to ovate Araucaria sp. “mature” foliage of
682 Zhou and Li (1994) and Shi et al. (2018) differs from A. pichileufensis and A. huncoensis
683 because the Antarctic leaves are significantly wider and broader toward the leaf apex.
684 Early Eocene Araucaria readiae from Regatta Point, Tasmania is based on a single leafy
685 branch specimen (Hill and Bigwood, 1987); A. readiae differs from both species described here
686 because the leaves are wider and have oblique stomatal orientation. Also from Tasmania, early
687 Eocene A. macrophylla and early Oligocene A. mollifolia both differ from the Patagonian species
688 described here in their significantly larger foliage (Hill et al., 2019). Similarly, early Oligocene
689 to late early Miocene A. balfourensis from Tasmania has wider leaves than either of the
690 Patagonian species (Hill et al., 2019).
691 Late Eocene-early Oligocene Araucarites alatisquamosus from the Loreto Formation,
692 Río de las Minas, Chile is based on ovuliferous complexes (Ohsawa et al., 2016). Araucarites
693 alatisquamosus differs from both fossil species described here from Argentina because the OCs
694 are significantly larger (> 12 mm longer than Araucaria pichileufensis and > 19 mm longer than
695 A. huncoensis). Both the OCs and seeds of Araucarites alatisquamosus are larger than any other
32
696 Cenozoic fossil attributed to Sect. Eutacta. The leaves from Río de las Minas are assigned to
697 Araucaria nathorstii Dusén (Sect. Araucaria), and an unnamed smaller leaf type is
698 morphologically similar to Sect. Eutacta, including oblique or perpendicular stomatal
699 orientation, with obtuse apices that are unlike the fossil leaves from either Laguna del Hunco or
700 Río Pichileufú.
701 Araucarites ruei from Port Jeanne d'Arc, Kerguelen Archipelago is based on dimorphic
702 foliage and OCs (Seward and Conway, 1934) and is probably lower Miocene in age (20–22 Ma;
703 Weis et al., 1993). The larger foliage form of Araucarites ruei has broad, triangular, imbricate
704 leaves with obtuse apices that differ from A. pichileufensis and A. huncoensis. The smaller leaf
705 form of Araucarites ruei also differs from the Patagonian fossils in having wide angles of
706 insertion and leaves well-spaced along the branch, similar to the foliage associated with A.
707 fildesensis. Varying widths of OCs are preserved for Araucarites ruei, showing differential
708 preservation of the lateral wings. The rounded apophysis region of Araucarites ruei and the
709 width of the OCs are more similar to A. huncoensis than A. pichileufensis but still distinct from
710 A. huncoensis.
711 Comparison with living Araucaria—The organs of both A. pichileufensis and A.
712 huncoensis consistently overlap or are within the size ranges of species from extant Sect. Eutacta
713 (Fig. 12; Farjon 2019, Mill et al. 2017). Both A. pichileufensis and A. huncoensis are on the small
714 end of leaf size for living Sect. Eutacta (Fig 12A; consistent with Merkhofer et al., 2015).
715 Araucaria pichileufensis has leaves similar in size to A. humboldtensis J. Buchholz, A.
716 heterophylla, and A. columnaris (Farjon, 2010). The average leaf size of A. huncoensis is closest
717 to the ranges of A. scopulorum de Laub. (Farjon, 2010). In extant specimens, a single leaf can be
718 extracted from the branch and measured in its entirety; however, the complete measurements of
33
719 an individual fossil leaf are usually obscured by overlapping adjacent leaves along the branch.
720 Based on extant collections, measurement of leaves along the spiral, imbricate leafy shoot fossil
721 results in an estimated 25-30% reduction compared with actual full leaf length.
722 Berry (1938) wrote that the leafy branches of A. pichileufensis most resembled either A.
723 heterophylla (Salisb.) Franco or A. columnaris (J.R. Forst.) Hook.. Florin (1940) found the leafy
724 branch morphology of A. pichileufensis to be most comparable with A. columnaris, A. subulata
725 Vieill., and A. cunninghamii. In form, we consider the lanceolate adpressed foliage of A.
726 pichileufensis to be comparable with the adult leaves of A. cunninghamii, A. heterophylla, and A.
727 subulata. The ovate imbricate form of Araucaria huncoensis is similar in shape to the adult
728 foliage of A. columnaris and A. humboldtensis.
729 Although both Patagonian species described here have small reproductive organs,
730 Araucaria huncoensis has some of the smallest ovuliferous complexes compared with the living
731 species of Eutacta and is closest in size to A. rulei F. Muell. (Fig. 12B). In general, extant
732 Araucaria are known to have larger male and female cones than those in the fossil record
733 (Gleiser et al., 2019), and the fossil species described here fit that pattern (Fig. 12B, C). The OCs
734 of A. pichileufensis are most similar to A. cunninghamii in size and in the flabellate shape.
735 With the caveat that fossil pollen cone sample size is very limited, the A. pichileufensis
736 specimen is similar in length to those from A. columnaris; however, the pollen cones of A.
737 columnaris are at least 10 mm wider than A. pichileufensis and have much larger
738 microsporophylls that are oblong, triangular, and spreading (Farjon, 2010). The width of the A.
739 pichileufensis pollen cones is most similar to A. cunninghamii, which has the narrowest extant
740 cones. A. cunninghamii cones have peltate, rhombic to rounded microsporophyll laminae and
741 thin, long basal bracts that are both similar to those of the A. pichileufensis specimens.
34
742 Similarly, Araucaria huncoensis pollen cones also have smaller widths similar to A.
743 cunninghamii or A. scopulorum, but the pollen cone lengths are within the ranges of A.
744 laubenfelsii Corbasson and A. nemorosa de Laub. (Farjon, 2010). However, the imbricate
745 microsporophylls of A. laubenfelsii are triangular, much larger than A. huncoensis, and have
746 obtuse apices; A. nemorosa also has much larger microsporophylls than A. huncoensis and ovate
747 laminae with denticulate margins (Farjon, 2010). Araucaria scopulorum pollen cones have
748 ovate-triangular laminae that can appear rhombic and resemble A. huncoensis, but the basal
749 bracts of A. scopulorum are much shorter and wider than those of A. huncoensis and highly
750 incurved at the apices. The basal bracts on the A. huncoensis specimens are much more slender
751 than basal bracts of A. scopulorum, appearing to taper distally instead of having an incurved
752 apex. Like A. pichileufensis, the triangular basal bracts of A. huncoensis are most similar to those
753 of A. cunninghamii.
754
755 RESULTS OF PHYLOGENETIC ANALYSIS
756 We obtained more than 200,000 most parsimonious trees of 15,569.592 steps in our total
757 evidence analysis. The relationships between Araucariaceae and the outgroup conifer families,
758 relationships within the agathioid clade, and the stem positions of the Jurassic to Late Cretaceous
759 Araucaria fossils all resolved identically to those previously reported (see Escapa and Catalano,
760 2013; Escapa et al., 2018). Figure 13 shows the strict consensus with bootstrap support values for
761 Araucariaceae, including the fossils Araucaria huncoensis and A. pichileufensis.
762 The fossil Araucaria pichileufensis resolved with strong support as sister to the well-
763 supported crown clade of Araucaria Section Eutacta, forming a stem lineage (Fig. 13), whereas
35
764 Araucaria huncoensis resolved with strong support within the crown of Sect. Eutacta. The total
765 group (crown + stem groups) of the Eutacta clade is supported morphologically by the mature
766 leaf-apex curvature (incurved; character 48; for the fossils, seen in the imbricate/adpressed
767 forms). The monophyly of the crown group of the Eutacta clade is supported by the mostly
768 perpendicular-oblique stomatal orientation on mature leaves (character 54).
769 Araucaria huncoensis is also placed as sister to the extant A. muelleri, within the New
770 Caledonian clade, but with very low support, as for the living species in the clade. In this
771 analysis, the New Caledonian clade is supported by three molecular characters and one
772 continuous character also not observable in the fossils, maximum tree height (character 9; ca. 34
773 m–ca. 53 m). The local grouping of A. muelleri and A. huncoensis is supported by the mostly
774 perpendicular stomatal orientation in mature leaves (character 54) and by pollen cone length
775 (character 4; 70–89 mm–109 mm), although these character states are not unique to this pair of
776 species, and the natural variation of cone length for A. huncoensis is unknown.
777
778 DISCUSSION
779 780 Phylogenetic relationships — The position of A. pichileufensis on the stem of Eutacta
781 (Fig. 13) is supported by the parallel orientation of its stomata (Figs. 6A, 6E, 6F,10D), which
782 appears to be plesiomorphic in Araucaria. The stomatal orientation distinguishes it from living
783 Sect. Eutacta species, which only have oblique or perpendicular stomata; parallel stomata are
784 seen in extant Sects. Araucaria, Bunya, and Intermedia (Stockey and Ko, 1986). Similarly, the
785 placement of A. huncoensis in the crown of Eutacta is supported by its perpendicular stomata.
36
786 However, the position of A. huncoensis in the New Caledonian clade of Sect. Eutacta, and the
787 grouping of A. muelleri and A. huncoensis within the New Caledonian clade, are not well
788 supported (see Results). Thus, we interpret our results to support placement of A. huncoensis in
789 the crown of Sect. Eutacta, which includes living species from Australia and New Guinea (A.
790 heterophylla and A. cunninghamii), but not in the New Caledonian clade.
791 Paleoecological and biogeographic implications — Araucaria Sect. Eutacta fits
792 biogeographically with the documented fossil assemblages from the Laguna del Hunco and Río
793 Pichileufú floras; many of the fossil genera present have nearest living relatives in present-day
794 subtropical and tropical montane rainforests of Australasia and southeast Asia (Appendix S1).
795 New Guinea montane forests are one of the closest modern analogs for Laguna del Hunco and
796 Río Pichileufú, with a large number of shared genera (Wilf et al., 2019) including living
797 Araucaria (with one species in Sect. Eutacta), Agathis, Dacrycarpus, Papuacedrus, Castanopsis,
798 Eucalyptus, Gymnostoma, and several other taxa from one or both of the two Eocene Patagonian
799 floras considered here (Brass, 1941; Soepadmo, 1972; Johns, 1989; Enright, 1995; Takeuchi,
800 1999; Ladiges et al., 2003). Thus, we infer that like many other taxa, Araucaria Section Eutacta
801 was a widespread component of Gondwanan forests (Table 1) and survived on Sahul (now
802 Australia and New Guinea). Eventually, the lineage dispersed to and radiated in New Caledonia.
803 Araucaria cunninghamii Aiton ex D. Don, considered the sister species to all other extant
804 Eutacta species (Setoguchi et al., 1998; Stefenon et al., 2006; Kranitz et al., 2014), is the most
805 dry-adapted of the Eutacta lineage, although it ranges into perhumid rainforests, especially in
806 New Guinea; in Australia, it is most common in dry vine forests and thickets with rainfall
807 between 630-1100 mm (Enright, 1995; Harden, 2006). The A. cunninghamii in New Guinea are
808 found in lower to upper montane forests at (90–) 500–1900 (–2800) m elevation, emerging from
37
809 a 30-meter canopy at heights of ca. 50–60 meters (Enright, 1995; Farjon, 2010). Recording
810 stations at about the same altitudes as the A. cunninghamii stands in New Guinea report average
811 annual rainfalls of ca. 1400–5700 mm (Gray, 1973) and mean annual temperatures of 19.0–
812 24.3°C (McAlpine et al., 1975). Species from the other shared genera mentioned between the
813 Patagonian fossil localities and New Guinea are found today in areas with mean annual
814 precipitation of ~1655–3285 mm and mean annual temperatures ranging from ~11–23°C
815 (Merkhofer et al. 2015; data include species from New Guinea, New Caledonia, Fiji, Malaysia,
816 New Zealand). Thus, Araucaria can tolerate a wider range of precipitation compared with many
817 of the other survivor genera. The New Guinea modern analog of the studied paleofloras, with
818 Araucaria Sect. Eutacta included in the list of fossil associations also found in the modern
819 assemblage, refines understanding of the Eocene paleoclimate and forest structure at Laguna del
820 Hunco and Río Pichileufú.
821 Previous analyses of Eocene vegetation of southern Argentina noted turnover related to
822 aridification and opening of forests during the middle Eocene, which is the age of the Río
823 Pichileufú flora (e.g., Barreda and Palazzesi, 2007; Palazzesi and Barreda, 2007; Dunn et al.,
824 2015; Zucol et al., 2018). Given the regional and global environmental and paleogeographic
825 changes that began during the middle Eocene and the drought tolerance of some extant
826 Araucaria Sect. Eutacta (Zimmer et al., 2016), it is possible that turnover in the Araucaria
827 species between Laguna del Hunco and Río Pichileufú, considered along with other floristic
828 evidence discussed earlier (Appendix S1), could be a signal of a biotic response to cooling and
829 drying. The lacustrine-volcanic depositional environment is similar between Laguna del Hunco
830 and Río Pichileufú (Aragón and Romero, 1984; Aragón and Mazzoni, 1997); thus, the Río
831 Pichileufú flora may record one of the final New Guinea-type rainforest environments in
38
832 Patagonia. During the early middle Eocene, Araucaria Sect. Eutacta populations remaining in
833 Patagonia would have been cut off from Antarctica to the south as the Drake Passage began to
834 deepen and widen (Lawver et al., 2011) and further restrained by dry, semiarid conditions still
835 present to the north (e.g., Melchor et al., 2002; Ziegler et al., 2003).
836
837 CONCLUSIONS
838 The confirmed presence of Araucaria Sect. Eutacta in early Eocene Patagonia just before
839 and during the initial isolation of South America adds to the Gondwanan biogeographic affinities
840 and southern connection of the Laguna del Hunco and Río Pichileufú floras. For the first time,
841 52.2 Ma Araucaria fossils from Laguna del Hunco, here described as Araucaria huncoensis sp.
842 nov., were found to be distinct from the fossil species Araucaria pichileufensis from the Río
843 Pichileufú locality, which is ca. 4.5 million years younger. This is the first detection of turnover
844 among the conifers at the two localities (except for the extremely rare Huncocladus laubenfelsii
845 from Laguna del Hunco), adding to the emerging evidence of floral turnover in Patagonia that
846 appears to be related to environmental change following the early Eocene climatic optimum and
847 the beginning of Antarctic separation.
848 Araucaria huncoensis and Araucaria pichileufensis are two of the most complete
849 representations of A. Sect. Eutacta known in the fossil record, with large sample sizes and multi-
850 organ preservation. Our new taxonomic treatment and combined evidence phylogenetic analysis
851 of these Araucaria fossil species both support their placement in Sect. Eutacta. Araucaria
852 pichileufensis was found in the stem of Eutacta, whereas A. huncoensis resolved within the
853 crown of Eutacta, although we caution that a single character, stomatal orientation, appears to
39
854 have been very influential at the corresponding node. The Patagonian fossils show that
855 Araucaria Sect. Eutacta was present in Patagonia 52.2 million years ago, before final
856 Gondwanan separation. Recent molecular divergence ages for the stem of Eutacta accommodate
857 the ages of these fossils, but the position of A. huncoensis suggests that the crown Eutacta
858 lineage evolved tens of millions of years before several molecular age estimates indicate. During
859 the initial isolation of South America, Sect. Eutacta survived in Patagonia and was thriving,
860 despite the climate changes and geographic restrictions that occurred after the early Eocene
861 climatic optimum.
862
863 ACKNOWLEDGMENTS
864 865 We thank M. Caffa, L. Canessa, R. Cúneo, M. Gandolfo, A. Iglesias, K. Johnson, L.
866 Reiner, E. Ruigomez, P. Puerta, S. Wing, J. Wingerath, and the many others involved in the
867 collection and curation of the Patagonian fossils. We thank A. Iglesias for leading the 2017 Río
868 Pichileufú field trip that produced a large increase in A. pichileufensis material from the type
869 locality. We also thank S. Chamberlain, P. Griffith, J. Tucker Lima, and the staff of the
870 Montgomery Botanical Center, Harvard Herbaria, and Huntington Library and Gardens for
871 assistance with and curation of the living material. We thank the editors and reviewers for their
872 comments, which have improved the manuscript. This research was supported by a Botanical
873 Society of America Graduate Student Research Grant; a Geological Society of America Graduate
874 Student Research Grant; a Penn State Geosciences Charles E. Knopf, Sr., Memorial Scholarship;
875 and a Penn State Geosciences Paul D. Krynine Scholarship (to GR-H); NSF grants DEB-
876 1556666, EAR-1925755, DEB-0919071, and DEB-0345750 (to PW and others); and National
40
877 Geographic Society grant 7337-02 (to PW and others). This research partially fulfilled
878 requirements for a 2019 M.S. in Geosciences at Pennsylvania State University for GR-H.
879
880 AUTHOR CONTRIBUTIONS
881 882 GR-H wrote the manuscript and collected and analyzed the data. PW participated in the
883 taxonomic description, provided funding for the project, and participated in nearly all the field
884 trips resulting in the new collections. IHE participated in the taxonomic and phylogenetic
885 analysis. AA-C participated in the study of the fossil cuticle and phylogenetic analysis. All
886 authors contributed feedback on drafts.
887
888 DATA AVAILABILITY
889
890 A high-resolution image archive of the fossils reported in this article is available open-access at
891 Figshare, doi: 10.6084/m9.figshare.11831124. The morphological character matrix is available at
892 MorphoBank, http://morphobank.org/permalink/?P3581.
893
894 SUPPORTING INFORMATION
895
896 Additional Supporting Information may be found online in the supporting information section at
897 the end of the article, Appendix S1: Taxa with modern systematics work, showing presence,
41
898 absence, and turnover at species level between early Eocene Laguna del Hunco and middle
899 Eocene Río Pichileufú, southern Argentina; Appendix S2: Living Araucaria Section Eutacta
900 collections vouchered at the Pennsylvania State University Herbarium (PAC), made by Gabriella
901 Rossetto-Harris at the Montgomery Botanical Center (MBC), Coral Gables, Florida, May 23-24,
902 2018; and Appendix S3: Combined matrix TNT file.
903
904 LITERATURE CITED
905 906 Andruchow‐ Colombo, A., I. H. Escapa, N. R. Cúneo, and M. A. Gandolfo. 2018. Araucaria
907 lefipanensis (Araucariaceae), a new species with dimorphic leaves from the Late
908 Cretaceous of Patagonia, Argentina. American Journal of Botany 105: 1067-1087.
909 Andruchow-Colombo, A., P. Wilf, and I. Escapa. 2019. A South American relative of
910 Phyllocladus? Huncocladus laubenfelsii gen. et sp. nov. (Podocarpaceae), from the early
911 Eocene of Laguna del Hunco, Patagonia, Argentina. Australian Systematic Botany 32:
912 290–309.
913 Aragón, E., and M. M. Mazzoni. 1997. Geología y estratigrafía del complejo volcánico
914 piroclástico del Río Chubut medio (Eoceno), Chubut, Argentina. Revista de la Asociación
915 Geológica Argentina 52: 243–256.
916 Aragón, E., and E. J. Romero. 1984. Geología, paleoambientes y paleobotánica de yacimientos
917 Terciarios del occidente de Río Negro, Neuquén y Chubut. Actas del IX Congreso
918 Geológico Argentino, San Carlos de Bariloche 4: 475–507.
919 Archangelsky, S. 1966. New gymnosperms from the Ticó flora, Santa Cruz Province, Argentina.
920 Bulletin of the British Museum (Natural History), Geology 13: 259–295.
42
921 Axsmith, B. J., I. H. Escapa, and P. Huber. 2008. An araucarian conifer bract-scale complex
922 from the lower Jurassic of Massachusetts: implications for estimating phylogenetic and
923 stratigraphic congruence in the Araucariaceae. Palaeontologia Electronica 11: article
924 11.13.13A.
925 Baldoni, A. M. 1979. Nuevos elementos paleofloristicos de la tafoflora de la Formación Spring
926 Hill, limite Jurasico-Cretacico subsuelo de Argentina y Chile austral. Ameghiniana 16:
927 103–119.
928 Barreda, V. D., and L. Palazzesi. 2007. Patagonian vegetation turnovers during the Paleogene-
929 early Neogene: origin of arid-adapted floras. Botanical Review 73: 31–50.
930 Barreda, V. D., L. Palazzesi, M. C. Tellería, L. Katinas, J. V. Crisci, K. Bremer, M. G. Passalia,
931 et al. 2010. Eocene Patagonia fossils of the daisy family. Science 329: 1621.
932 Barreda, V.D., M. C. Zamaloa, M. A. Gandolfo, C. Jaramillo, and P. Wilf. 2020. Early Eocene
933 spore and pollen assemblages from the Laguna del Hunco fossil-lake beds, Patagonia,
934 Argentina. In press, International Journal of Plant Sciences.
935 Berry, E. W. 1925. A Miocene flora from Patagonia. Johns Hopkins University Studies in
936 Geology 6: 183–251.
937 Berry, E. W. 1938. Tertiary flora from the Río Pichileufú, Argentina. Geological Society of
938 America Special Paper 12: 1–149.
939 Biffin, E., R. S. Hill, and A. J. Lowe. 2010. Did Kauri (Agathis: Araucariaceae) really survive
940 the Oligocene drowning of New Zealand? Systematic Biology 59: 594–602.
941 Bigwood, A. J., and R. S. Hill. 1985. Tertiary Araucarian macrofossils from Tasmania.
942 Australian Journal of Botany 33: 645–656.
43
943 Bijl, P. K., J. a. P. Bendle, S. M. Bohaty, J. Pross, S. Schouten, L. Tauxe, C. E. Stickley, et al.
944 2013. Eocene cooling linked to early flow across the Tasmanian Gateway. Proceedings of
945 the National Academy of Sciences USA 110: 9645–9650.
946 Bose, M. N., and H. K. Maheshwari. 1973. Some detached seed-scales belonging to
947 Araucariaceae from the Mesozoic rocks of India. Geophytology 3: 205–214.
948 Brass, L. 1941. The 1938-1939 expedition to the Snow Mountains, Netherlands New Guinea.
949 Journal of the Arnold Arboretum 22: 271–342.
950 Cantrill, D. J. 1992. Araucarian foliage from the Lower Cretaceous of Southern Victoria,
951 Australia. International Journal of Plant Sciences 153: 622–645.
952 Cantrill, D. J., and H. J. Falcon-Lang. 2001. Cretaceous (Late Albian) coniferales of Alexander
953 Island, Antarctica. 2. Leaves, reproductive structures and roots. Review of Palaeobotany
954 and Palynology 115: 119–145.
955 Chambers, T. C., A. N. Drinnan, and S. McLoughlin. 1998. Some morphological features of
956 Wollemi pine (Wollemia nobilis: Araucariaceae) and their comparison to Cretaceous
957 plant fossils. International Journal of Plant Sciences 159: 160–171.
958 Cookson, I. C., and S. L. Duigan. 1951. Tertiary Araucariaceae from south-eastern Australia,
959 with notes on living species. Australian Journal of Scientific Research Series B,
960 Biological Sciences 4: 415–449.
961 Crisp, M. D., and L. G. Cook. 2011. Cenozoic extinctions account for the low diversity of extant
962 gymnosperms compared with angiosperms. New Phytologist 192: 997–1009.
963 de Laubenfels, D. J. 1953. The external morphology of coniferous leaves. Phytomorphology 3:
964 1–20.
44
965 de Laubenfels, D. J. 1972. Flore de la Nouvelle Calédonie et dépendances, 1–168. Muséum
966 National D'Histoire Naturelle, Paris.
967 de Laubenfels, D. J. 1988. Coniferales. Flora Malesiana series I 10: 337–453.
968 Dunn, R. E., C. A. E. Strömberg, R. H. Madden, M. J. Kohn, and A. A. Carlini. 2015. Linked
969 canopy, climate, and faunal change in the Cenozoic of Patagonia. Science 347: 258–261.
970 Endlicher, S. 1847. Synopsis coniferarum, 1–368. Scheitlin und Zollikofer, Sangalli (Sankt
971 Gallen), Switzerland.
972 Enright, N. J. 1995. Conifers of tropical Australasia. In N. J. Enright and R. S. Hill [eds.],
973 Ecology of the southern conifers, 197–222. Smithsonian Institution Press, Washington,
974 D.C.
975 Escapa, I. H., and S. A. Catalano. 2013. Phylogenetic analysis of Araucariaceae: integrating
976 molecules, morphology, and fossils. International Journal of Plant Sciences 174: 1153–
977 1170.
978 Escapa, I. H., A. Iglesias, P. Wilf, S. A. Catalano, M. A. Caraballo‐ Ortiz, and N. R. Cúneo.
979 2018. Agathis trees of Patagonia's Cretaceous‐ Paleogene death landscapes and their
980 evolutionary significance. American Journal of Botany 105: 1345–1368.
981 Escapa, I. H., and D. Pol. 2011. Dealing with incompleteness: new advances for the use of fossils
982 in phylogenetic analysis. Palaios 26: 121–124.
983 Farjon, A. 2010. A handbook of the world's conifers. Brill, Leiden, Netherlands.
984 Florin, R. 1931. Untersuchungen zur Stammesgeschichte der Coniferales und Cordaitales.
985 Kungliga Svenska Vetenskapsakademiens Handlingar 10: 1–588.
986 Florin, R. 1940. The Tertiary fossil conifers of south Chile and their phytogeographical
987 significance. Kungliga Svenska Vetenskapsakademiens Handlingar 19: 1–107.
45
988 Gandolfo, M. A., E. J. Hermsen, M. C. Zamaloa, K. C. Nixon, C. C. González, P. Wilf, N. R.
989 Cúneo, and K. R. Johnson. 2011. Oldest known Eucalyptus macrofossils are from South
990 America. PLoS One 6: e21084.
991 Gaudeul, M., G. Rouhan, M. F. Gardner, and P. M. Hollingsworth. 2012. AFLP markers provide
992 insights into the evolutionary relationships and diversification of New Caledonian
993 Araucaria species (Araucariaceae). American Journal of Botany 99: 68–81.
994 Gilmore, S., and K. D. Hill. 1997. Relationships of the Wollemi pine (Wollemia nobilis) and a
995 molecular phylogeny of the Araucariaceae. Telopea 7: 275–291.
996 Gleiser, G., K. L. Speziale, S. A. Lambertucci, F. Hiraldo, J. L. Tella, and M. A. Aizen. 2019.
997 Uncoupled evolution of male and female cone sizes in an ancient conifer lineage.
998 International Journal of Plant Sciences 180: 72–80.
999 Goloboff, P. A., and S. A. Catalano. 2016. TNT version 1.5, including a full implementation of
1000 phylogenetic morphometrics. Cladistics 32: 221–238.
1001 Goloboff, P., J. Farris, and K. Nixon. 2003. TNT: Tree Analysis Using New Technology. Ver.
1002 1.0. Program and documentation freely available: https://cladistics.org/tnt/.
1003 Goloboff, P. A., J. S. Farris, and K. C. Nixon. 2008. TNT, a free program for phylogenetic
1004 analysis. Cladistics 24: 774–786.
1005 Gradstein, F. M., J. G. Ogg, M. Schmitz, and G. Ogg. 2012. The geologic time scale 2012, 1–
1006 1144. Elsevier, Amsterdam, Netherlands.
1007 Gray, B. 1973. Distribution of Araucaria in Papua New Guinea. Research Bulletin No. 1, Papua
1008 New Guinea Department of Forestry, Port Moresby, Papua New Guinea.
1009 Harden, G. J. M., W.J.F.; Williams, John B. 2006. Rainforest trees and shrubs: a field guide to
1010 their identification. Gwen Harden Publishing, Nambucca Heads, Australia.
46
1011 Harris, T. M. 1979. The Yorkshire Jurassic Flora V. Coniferales. Trustees of the British
1012 Museum, London, UK.
1013 Hill, R. S. 1995. Conifer origin, evolution, and diversification in the Southern Hemisphere. In N.
1014 J. Enright and R. S. Hill [eds.], Ecology of the southern conifers, 10–29. Smithsonian
1015 Institution Press, Washington, D.C.
1016 Hill, R. S., and A. J. Bigwood. 1987. Tertiary gymnosperms from Tasmania: Araucariaceae.
1017 Alcheringa 11: 325–335.
1018 Hill, R. S., and T. J. Brodribb. 1999. Southern conifers in time and space. Australian Journal of
1019 Botany 47: 639–696.
1020 Hill, R. S., G. J. Jordan, R. J. Carpenter, and R. Paull. 2019. Araucaria Section Eutacta
1021 macrofossils from the Cenozoic of Southeastern Australia. International Journal of Plant
1022 Sciences 180: 902–921.
1023 Hollis, C. J., K. W. R. Taylor, L. Handley, R. D. Pancost, M. Huber, J. B. Creech, B. R. Hines, et
1024 al. 2012. Early Paleogene temperature history of the Southwest Pacific Ocean:
1025 reconciling proxies and models. Earth and Planetary Science Letters 349–350: 53–66.
1026 Iannelli, S. B., V. D. Litvak, L. Fernández Paz, A. Folguera, M. E. Ramos, and V. A. Ramos.
1027 2017. Evolution of Eocene to Oligocene arc-related volcanism in the North Patagonian
1028 Andes (39–41°S), prior to the break-up of the Farallon plate. Tectonophysics 696–697:
1029 70–87.
1030 Johns, R. J. 1989. The influence of drought on tropical rainforest vegetation in Papua New
1031 Guinea. Mountain Research and Development 9: 248–251.
1032 Jones, W. G., K. D. Hill, and J. M. Allen. 1995. Wollemia nobilis, a new living Australian genus
1033 and species in the Araucariaceae. Telopea 6: 173–176.
47
1034 Kendall, M. W. 1949. A Jurassic member of the Araucariaceae. Annals of Botany 13: 151–161.
1035 Kershaw, P., and B. Wagstaff. 2001. The southern conifer family Araucariaceae: history, status,
1036 and value for paleoenvironmental reconstruction. Annual Review of Ecology and
1037 Systematics 32: 397–414.
1038 Kooyman, R. M., P. Wilf, V. D. Barreda, R. J. Carpenter, G. J. Jordan, J. M. K. Sniderman, A.
1039 Allen, et al. 2014. Paleo-Antarctic rainforest into the modern Old World Tropics: the rich
1040 past and threatened future of the "southern wet forest survivors". American Journal of
1041 Botany 101: 2121–2135.
1042 Kranitz, M. L., E. Biffin, A. Clark, M. L. Hollingsworth, M. Ruhsam, M. F. Gardner, P. Thomas,
1043 et al. 2014. Evolutionary diversification of New Caledonian Araucaria. PLoS One 9:
1044 e110308.
1045 Kunzmann, L. 2007. Araucariaceae (Pinopsida): aspects in palaeobiogeography and
1046 palaeobiodiversity in the Mesozoic. Zoologischer Anzeiger 246: 257–277.
1047 Ladiges, P. Y., F. Udovicic, and G. Nelson. 2003. Australian biogeographical connections and
1048 the phylogeny of large genera in the plant family Myrtaceae. Journal of Biogeography
1049 30: 989–998.
1050 Lauretano, V., J. C. Zachos, and L. J. Lourens. 2018. Orbitally paced carbon and deep‐ sea
1051 temperature changes at the peak of the Early Eocene Climatic Optimum.
1052 Paleoceanography and Paleoclimatology 33: 1050–1065.
1053 Lawver, L. A., L. M. Gahagan, and I. W. D. Dalziel. 2011. A different look at gateways: Drake
1054 Passage and Australia/Antarctica. In J. B. Anderson and J. S. Wellner [eds.], Tectonic,
1055 climatic, and cryospheric evolution of the Antarctic Peninsula, 5-33. AGU, Washington,
1056 DC.
48
1057 Leslie, A.B., J. M. Beaulieu, G. Holman, C. S. Campbell, W. Mei, L. R. Raubeson, and S.
1058 Mathews. 2018. An overview of extant conifer evolution from the perspective of the
1059 fossil record. American Journal of Botany 105:1531-1544.
1060 Leslie, A. B., J. M. Beaulieu, H. S. Rai, P. R. Crane, M. J. Donoghue, and S. Mathews. 2012.
1061 Hemisphere-scale differences in conifer evolutionary dynamics. Proceedings of the
1062 National Academy of Sciences USA 109: 16217–16221.
1063 Liu, N., Y. Zhu, Z. Wei, J. Chen, Q. Wang, S. Jian, D. Zhou, et al. 2009. Phylogenetic
1064 relationships and divergence times of the family Araucariaceae based on the DNA
1065 sequences of eight genes. Chinese Science Bulletin 54: 2648–2655.
1066 McAlpine, J. R., G. Keig, and K. Short. 1975. Climatic tables for Papua New Guinea. Division
1067 of Land Use Research Technical Paper No. 37, Commonwealth Scientific and Industrial
1068 Research Organization, Melbourne, Australia.
1069 McLoughlin, S. 2001. The breakup history of Gondwana and its impact on pre-Cenozoic floristic
1070 provincialism. Australian Journal of Botany 49: 271–300.
1071 Melchor, R. N., J. F. Genise, and S. E. Miquel. 2002. Ichnology, sedimentology and
1072 paleontology of Eocene calcareous paleosols from a palustrine sequence, Argentina.
1073 Palaios 17: 16–35.
1074 Merkhofer, L., P. Wilf, M. T. Haas, R. M. Kooyman, L. Sack, C. Scoffoni, and N. R. Cúneo.
1075 2015. Resolving Australian analogs for an Eocene Patagonian paleorainforest using leaf
1076 size and floristics. American Journal of Botany 102: 1160–1173.
1077 Mill, R., M. Ruhsam, P. Thomas, M. Gardner, and P. Hollingsworth. 2017. Araucaria goroensis
1078 (Araucariaceae), a new monkey puzzle from New Caledonia, and nomenclatural notes on
1079 Araucaria muelleri. Edinburgh Journal of Botany 74: 123–139.
49
1080 Ohsawa, T. A., A. Yabe, T. Yamada, K. Uemura, K. Terada, M. Leppe, L. F. Hinojosa, and H.
1081 Nishida. 2016. Araucarian leaves and cone scales from the Loreto Formation of Río de
1082 Las Minas, Magellan Region, Chile. Botany 94: 805–815.
1083 O'Leary, M. A., and S. G. Kaufman. 2012. MorphoBank 3.0: Web application for morphological
1084 phylogenetics and taxonomy. http://www.morphobank.org.
1085 Palazzesi, L., and V. Barreda. 2007. Major vegetation trends in the Tertiary of Patagonia
1086 (Argentina): a qualitative paleoclimatic approach based on palynological evidence. Flora
1087 202: 328–337.
1088 Panti, C., R. R. Pujana, M. C. Zamaloa, and E. J. Romero. 2012. Araucariaceae macrofossil
1089 record from South America and Antarctica. Alcheringa 36: 1–22.
1090 Pearson, P. N., B. E. Van Dongen, C. J. Nicholas, R. D. Pancost, S. Schouten, J. M. Singano, and
1091 B. S. Wade. 2007. Stable warm tropical climate through the Eocene Epoch. Geology 35:
1092 211–214.
1093 Petersen, C. S. 1946. Estudios geológicos en la región del Río Chubut medio. Dirección de
1094 Minas y Geología Boletín 59: 1–137.
1095 Pole, M. S. 1992. Early Miocene flora of the Manuherikia Group, New Zealand. 2. Conifers.
1096 Journal of the Royal Society of New Zealand 22: 287–302.
1097 Pole, M. S. 1995. Late Cretaceous macrofloras of eastern Otago, New Zealand: Gymnosperms.
1098 Australian Systematic Botany 8: 1067–1106.
1099 Pole, M. S. 2008. The record of Araucariaceae macrofossils in New Zealand. Alcheringa 32:
1100 405–426.
50
1101 Ruhsam, M., H. S. Rai, S. Mathews, T. G. Ross, S. W. Graham, L. A. Raubeson, W. Mei, et al.
1102 2015. Does complete plastid genome sequencing improve species discrimination and
1103 phylogenetic resolution in Araucaria? Molecular Ecology Resources 15: 1067–1078.
1104 Setoguchi, H., T. A. Osawa, J.-C. Pintaud, T. Jaffré, and J.-M. Veillon. 1998. Phylogenetic
1105 relationships within Araucariaceae based on rbcL gene sequences. American Journal of
1106 Botany 85: 1507–1516.
1107 Seward, A. C. 1903. Fossil floras of Cape Colony. Annals of the South African Museum 4: 1–
1108 122.
1109 Seward, A. C., and V. Conway. 1934. A phytogeographical problem: fossil plants from the
1110 Kerguelen Archipelago. Annals of Botany 48: 715–741.
1111 Shi, G., H. Li, A. B. Leslie, and Z. Zhou. 2018. Araucaria bract-scale complex and associated
1112 foliage from the early-middle Eocene of Antarctica and their implications for Gondwanan
1113 biogeography. Historical Biology doi: 10.1080/08912963.2018.1472255.
1114 Soepadmo, E. 1972. Fagaceae. Flora Malesiana - Series 1, Spermatophyta 7: 265–403.
1115 Spitzer, M., Wildenhain, J., Rappsilber, J., and M. Tyers. 2014. BoxPlotR: a web tool for
1116 generation of box plots. Nature Methods 11: 121–122. Tool accessed 10/8/19 at
1117 http://shiny.chemgrid.org/boxplotr/.
1118 Stefanovic, S., M. Jager, J. Deutsch, J. Broutin, and M. Masselot. 1998. Phylogenetic
1119 relationships of conifers inferred from partial 28S rRNA gene sequences. American
1120 Journal of Botany 85: 688–697.
1121 Stefenon, V., O. Gailing, and R. Finkeldey. 2006. Phylogenetic relationship within genus
1122 Araucaria (Araucariaceae) assessed by means of AFLP fingerprints. Silvae Genetica 55:
1123 45–52.
51
1124 Stockey, R. A. 1982. The Araucariaceae: an evolutionary perspective. Review of Palaeobotany
1125 and Palynology 37: 133–154.
1126 Stockey, R. A. 1994. Mesozoic Araucariaceae: morphology and systematic relationships.
1127 Journal of Plant Research 107: 493–502.
1128 Stockey, R. A., and H. Ko. 1986. Cuticle micromorphology of Araucaria de Jussieu. Botanical
1129 Gazette 147: 508–548.
1130 Stockey, R. A., and T. N. Taylor. 1978. Cuticular features and epidermal patterns in the genus
1131 Araucaria de Jussieu. Botanical Gazette 139: 490–498.
1132 Takeuchi, T. 1999. New plants from Crater Mt., Papua New Guinea, and an annotated checklist
1133 of the species. Sida 18: 941–986.
1134 Troncoso, A., and E. J. Romero. 1993. Consideraciones acerca de las coníferas del Mioceno de
1135 Chile central occidental. Boletín del Museo Nacional de Historia Natural, Chile 44: 47–
1136 71.
1137 van der Ham, R. W. J. M., J. W. M. Jagt, S. Renkens, and J. Van Konijnenburg-Van Cittert.
1138 2010. Seed-cone scales from the upper Maastrichtian document the last occurrence in
1139 Europe of the Southern Hemisphere conifer family Araucariaceae. Palaeogeography,
1140 Palaeoclimatology, Palaeoecology 291: 469–473.
1141 Weis, D., F. Frey, H. Leyrit, and I. Gautier. 1993. Kerguelen Archipelago revisited: geochemical
1142 and isotopic study of the Southeast Province lavas. Earth and Planetary Science Letters
1143 118: 101-119.
1144 Wilde, M. H., and A. J. Eames. 1952. The ovule and 'seed' of Araucaria bidwilli with discussion
1145 of the taxonomy of the genus. II. Taxonomy. Annals of Botany 16: 28–49.
52
1146 Wilf, P. 2012. Rainforest conifers of Eocene Patagonia: attached cones and foliage of the extant
1147 Southeast Asian and Australasian genus Dacrycarpus (Podocarpaceae). American
1148 Journal of Botany 99: 562–584.
1149 Wilf, P., M. R. Carvalho, M. A. Gandolfo, and N. R. Cúneo. 2017b. Eocene lantern fruits from
1150 Gondwanan Patagonia and the early origins of Solanaceae. Science 355: 71–75.
1151 Wilf, P., N. R. Cúneo, I. H. Escapa, D. Pol, and M. O. Woodburne. 2013. Splendid and seldom
1152 isolated: the paleobiogeography of Patagonia. Annual Review of Earth and Planetary
1153 Sciences 41: 561–603.
1154 Wilf, P., N. R. Cúneo, K. R. Johnson, J. F. Hicks, S. L. Wing, and J. D. Obradovich. 2003. High
1155 plant diversity in Eocene South America: evidence from Patagonia. Science 300: 122–
1156 125.
1157 Wilf, P., M. P. Donovan, N. R. Cúneo, and M. A. Gandolfo. 2017a. The fossil flip-leaves
1158 (Retrophyllum, Podocarpaceae) of southern South America. American Journal of Botany
1159 104: 1344–1369.
1160 Wilf, P., and I. H. Escapa. 2015. Green Web or megabiased clock? Patagonian plant fossils speak
1161 on evolutionary radiations. New Phytologist 207: 283–290.
1162 Wilf, P., I. H. Escapa, N. R. Cúneo, R. M. Kooyman, K. R. Johnson, and A. Iglesias. 2014. First
1163 South American Agathis (Araucariaceae), Eocene of Patagonia. American Journal of
1164 Botany 101: 156–179.
1165 Wilf, P., K. R. Johnson, N. R. Cúneo, M. E. Smith, B. S. Singer, and M. A. Gandolfo. 2005.
1166 Eocene plant diversity at Laguna del Hunco and Río Pichileufú, Patagonia, Argentina.
1167 American Naturalist 165: 634–650.
53
1168 Wilf, P., S. A. Little, A. Iglesias, M. C. Zamaloa, M. A. Gandolfo, N. R. Cúneo, and K. R.
1169 Johnson. 2009. Papuacedrus (Cupressaceae) in Eocene Patagonia: a new fossil link to
1170 Australasian rainforests. American Journal of Botany 96: 2031–2047.
1171 Wilf, P., K. C. Nixon, M. A. Gandolfo, and N. R. Cúneo. 2019. Eocene Fagaceae from Patagonia
1172 and Gondwanan legacy in Asian rainforests. Science 364: eaaw5139.
1173 Wilf, P., B. S. Singer, M. C. Zamaloa, K. R. Johnson, and N. R. Cúneo. 2010. Early Eocene
1174 40Ar/39Ar age for the Pampa de Jones plant, frog, and insect biota (Huitrera Formation,
1175 Neuquén Province, Patagonia, Argentina). Ameghiniana 47: 207–216.
1176 Wilkinson, M. 2003. Missing entries and multiple trees: instability, relationships, and support in
1177 parsimony analysis. Journal of Vertebrate Paleontology 23: 311–323.
1178 Zachos, J., M. Pagani, L. Sloan, E. Thomas, and K. Billups. 2001. Trends, rhythms, and
1179 aberrations in global climate 65 Ma to present. Science 292: 686–693.
1180 Zhou, Z., and H. Li. 1994. Early Tertiary gymnosperms from Fildes Peninsula, King George
1181 Island, Antarctica. In Y. Shen [ed.], Stratigraphy and paleontology of Fildes Peninsula,
1182 King George Island, Antarctica. State Antarctic Committee, Monographs 3, 191–221.
1183 Science Press, Beijing.
1184 Ziegler, A. M., G. Eshel, P. M. Rees, T. A. Rothfus, D. B. Rowley, and D. Sunderlin. 2003.
1185 Tracing the tropics across land and sea: Permian to present. Lethaia 36: 227–254.
1186 Zimmer, H. C., T. J. Brodribb, S. Delzon, and P.J. Baker. 2016. Drought avoidance and
1187 vulnerability in the Australian Araucariaceae. Tree Physiology 36: 218–228.
1188 Zucol, A. F., J. M. Krause, M. Brea, M. S. Raigemborn, and S. D. Matheos. 2018. Emergence of
1189 grassy habitats during the greenhouse–icehouse systems transition in the Middle Eocene
1190 of southern South America. Ameghiniana 55: 451–482.
54 Table 1. Cenozoic fossil occurrences of Araucaria Section Eutacta and measurements.
Species Age (radioisotopic age Provenance Citation OC length OC width Seed length Seed width Leaf length Leaf width Pollen cone Pollen cone of associated rocks, (mm) (mm) (mm) (mm) (mm) (mm) length width (mm) Ma) (mm) Araucaria pichileufensis E.W. middle Eocene (47.74 ± Río Pichileufú, Río Berry 1938; this 20.5–26.8 23.4–33.4 11.5–17.9 4.3–7.3 2.9–9.4 0.9–2.4 55 7.4 Berry 0.05 40Ar/39Ar) Negro, Argentina study
Araucaria huncoensis Rossetto- early Eocene (52.22 ± Laguna del Hunco, This study 13.3–24.5 13.9–24.9 8.5–17 3.2–8.8 2.4–8.2 1.1–3.0 86 8 Harris, sp. nov. 0.22 40Ar/39Ar) Chubut, Argentina
A. cf. pichileufensis E.W. Berry early Eocene (54.24 ± Pampa de Jones, Neuquén, Wilf et al. 2010 16.9 15 10.8 4.5 ? ? ? ? 0.45 40Ar/39Ar) Argentina
A. readiae Hill & Bigwood early Eocene Regatta Point, Tasmania, Hill and Bigwood ? ? ? ? 6–7 3–4 ? ? Australia 1987; Hill 1990
A. macrophylla R.S. Hill, G.J. early Eocene Regatta Point and Lowana Hill et al. 2019 ? ? ? ? 16–37 9.5–10.5 ? ? Jordan, R.J. Carpenter et R. Road, Tasmania, Paull Australia A. fildesensis Shi, Li, Leslie et early - middle Eocene Fildes Peninsula, King Shi et al. 2018 22 17 16 4.8 - - ? ? Zhou (52 ± 1–43 ± 2 K/Ar George Island, West and Rb/Sr) Antarctica Araucaria sp. early - middle Eocene Fildes Peninsula, King Zhou and Li 1994; Shi - - - - 10–14 1 ? ? (53 ± 1–43 ± 2 K/Ar George Island, West et al. 2018 and Rb/Sr) Antarctica Araucarites ruei Seward et early Miocene Port Jeanne d'Arc, Seward and Conway ? 15 ? ? 12 7 ? ? Conway Kerguelen Archipelago, 1934 French Southern and Antarctic Lands Araucarites alatisquamosus late Eocene - early Río de las Minas, Punta Ohsawa et al. 2016 32–33 30–34 18–24 5–5.5 - - ? ? Ohsawa et H. Nishida Oligocene (36.48 ± Arenas, Chile 0.47 U-Th-Pb) Araucarites sp. late Eocene- early Río de las Minas, Punta Ohsawa et al. 2016 12 12 7.6 1.6 - - ? ? Oligocene (36.48 ± Arenas, Chile 0.47 U-Th-Pb) Araucaria sp. late Eocene- early Río de las Minas, Punta Ohsawa et al. 2016 - - - - 11–13 6–7 ? ? Oligocene (36.48 ± Arenas, Chile 0.47 U-Th-Pb) A. mollifolia R.S. Hill, G.J. early Oligocene Lea River, Tasmania, Hill et al. 2019 ? ? ? ? 20–21 9–10.5 ? ? Jordan, R.J. Carpenter et R. Australia Paull A. ligniticii Cookson et Duigan Oligocene Yallourn, Victoria, Cookson and Duigan 11–18 9–11 ? ? 3–8 1–1.5 20* 5 Australia 1951; Hill 1990
A. balfourensis R.S. Hill, G.J. early Oligocene-late Little Rapid River, Hill et al. 2019 ? ? ? ? 7–10 4.5–5.5 ? ? Jordan, R.J. Carpenter et R. early Miocene Balfour, and Pioneer, Paull Tasmania, Australia A. uncinatus Hill late Oligocene-early Monpeelyata, Tasmania, Hill 1990 ? ? ? ? 5 2 ? ? Miocene Australia
A. planus Hill late Oligocene-early Monpeelyata, Tasmania, Hill 1990; Hill et al., ? ? ? ? 4 2 ? ? Miocene Australia 2019
A. prominens Hill late Oligocene-early Monpeelyata, Tasmania, Hill 1990 ? ? ? ? 5 > 2 ? ? Miocene Australia
Araucaria sp. Miocene Bannockburn, New Pole 1992 25–30 20–25 ? ? 4.6–7 0.8–2.4 35* 9 Zealand
Araucaria sp. 2 Miocene Boca Pupuya, Matanzas, Troncoso and Romero ? ? ? ? 5 1.5–2.5 ? ? Chile 1993 Notes: OC= ovuliferous complex; *=incomplete cone measurement.
55 FIGURE CAPTIONS
Figure 1. Eocene fossil localities Río Pichileufú (RP) and Laguna del Hunco (LH) on a paleoreconstruction of
Patagonia and the tip of the Antarctic Peninsula. Modern coastlines are outlined in black. From a paleoglobe for
the early Eocene (Ypresian), 56 million years ago, by C. R. Scotese, PALEOMAP Project.
Figure 2. Material of extant Araucaria Section Eutacta species for comparison with fossils. A. A. cunninghamii
juvenile leafy branch, showing linear leaves with pointed apices. MBC, Coral Gables, Florida, PAC 107296. B.
A. cunninghamii leafy branch, intermediate foliage, showing continuous variation in leaf size and shape MBC,
PAC 107296. C. A. cunninghamii mature foliage characteristic shoot apex (a) consisting of a cluster of small,
incompletely formed leaves. Also shows terminal attachment of male cones and thin microsporophyll stalks
(see inset). Morobe Province, Papua New Guinea, A, Coll. J. Havel 7549. D. Dense leafy branch litter under A.
luxurians at MBC. E. A. cunninghamii pollen cones showing helical arrangement of microsporophylls with
pointed laminae apices and sharp-pointed, elongate, triangular subtending basal bracts retained along the cone
body. Queensland, Australia, A, Coll. F. H. Weatherhead, s.n. F. A. cunninghamii, detail of juvenile leaf form,
showing broad attachment at the base, keeling on both leaf surfaces, and a sharp apex. MBC, PAC 107296. G.
A. cunninghamii, detail of intermediate leaf form, showing leaves thickened, incurved, and tapering toward the
pointed apex. MBC, PAC 107296. H. A. cunninghamii, detail of mature leaf form, showing wider, more
imbricated leaves compared to the intermediate leaf form that are strongly keeled abaxially, and softly- to non-
keeled adaxially. MBC, PAC 107296. I. A. cunninghamii, cross-section of seed cone showing overlapping,
helical arrangement of individual ovuliferous complexes (OC) and their projecting spiny apices. Queensland,
Australia, A, Coll. F. H. Weatherhead, s.n. J. A. cunninghamii, individual OCs from a shattered cone. The
central bodies of the OCs have a single central embedded seed (s), triangular ligule (lig) above the seed, apophysis (aph), and curved spine (sp) that projects distally from the apophysis. Papua Province, Western New
Guinea, L.J. Brass and C. Verateagh 11175. K. A. cunninghamii, detail of the triangular ligule (lig) on the OC,
56 projecting above the seed apex and into the apophysis (aph) region. The spine (sp) is curved toward the ligule.
MBC, PAC 107296.
Figure 3. Selected foliage of Araucaria pichileufensis E.W. Berry from Río Pichileufú. A, B: Previously
unfigured syntypes from the Berry 1938 type collection. A. USNM 545222, subulate spreading leaf form. B.
USNM 545210, continuous variation between adpressed and subulate spreading leaves. C. BAR 5581,
continuous variation between more spreading, longer leaves and shorter more adpressed leaves. D. BAR 5608,
keeled leaves spreading from the branch, then curving around the branch toward the apex. E. BAR 5576b, leaf
bases and subulate spreading foliage. F. BAR 5592a, continuous variation in lanceolate adpressed foliage, with
sections toward the apex of the branch that are less adpressed. G, H: Syntypes from the Berry 1938 collection.
G. USNM 40383a, lanceolate adpressed leaf form. H. USNM 40383d, adpressed foliage with a secondary
episode of growth near the apex. I. BAR 5602, lanceolate adpressed leaf form. J. BAR 5347, the most
adpressed and imbricate leafy branch recovered from Río Pichileufú.
Figure 4. Selected foliage of Araucaria huncoensis sp. nov. from Laguna del Hunco. A. MPEF-Pb 10593,
acicular falcate form with central axis of the branch visible. B. MPEF-Pb 10594, acicular form with an episode
of secondary growth near the apex. C. MPEF-Pb 10597, acicular falcate foliage with irregular leaf spacing. D.
MPEF-Pb 10596, acicular falcate form. E. MPEF-Pb 10585, ovate imbricate form with an episode of secondary growth emerging near the base of the preserved shoot. F. MPEF-Pb 10603b, imbricate ovate form showing full
adaxial leaves and leaf bases. G. MPEF-Pb 10580, showing central axis of the branch and short, broad,
incurving leaves in lateral view. H. MPEF-Pb 10605, ovate imbricate form showing an episode of growth and a
branching point at the apex of the shoot. I. MPEF-Pb 10602, ovate imbricate form, showing the adaxial surface
57 of the leaves along the branch. J. MPEF-Pb 10613 ovate imbricate form with at least four branchlets emerging
from the apex.
Figure 5. Fossil foliage details of Araucaria pichileufensis (A, B, E, F) and Araucaria huncoensis sp. nov. (C,
D, G–I). A.. A pichileufensis subulate spreading form, showing awl-shaped, spreading foliage with keeled
abaxial profiles. BAR 5612. B. A. pichileufensis lanceolate adpressed form, showing the abaxial face and
incurving lateral profile of leaves. BAR 5576b. C. A. huncoensis acicular falcate form, showing tightly packed,
incurving, acicular, slender leaves. MPEF-Pb 10593. D. A. huncoensis ovate imbricate form, showing broad,
ovate abaxial face and incurving lateral profile of leaves. MPEF-Pb 10603a. E. A. pichileufensis leaf bases, showing the broad attachment of the leaves where the leaf has completely broken off. BAR 5579a. F. A.
pichileufensis leafy branch apex, showing small, incompletely formed leaves coming together into a rounded
apex. BAR4225. G. A. huncoensis, ovate imbricate form showing a branching point toward the base of the
specimen and an additional growth episode emerging from the right branch. MPEF-Pb 10606. H. A. huncoensis
leaf bases, showing broad attachment of the leaves where the leaf has broken off. MPEF-Pb 10584. I. A.
huncoensis leafy branch apex, showing small leaves forming a cluster at the apex. MPEF-Pb 10589.
Figure 6. In situ fossil leaf cuticle of Araucaria pichileufensis (A, E, F) and Araucaria huncoensis sp. nov. (B–
D, G, H). A.. A pichileufensis, profile of leaves under epifluorescence. A darker ridge (m) is the leaf margin,
showing the separation of the abaxial (ab) and adaxial (ad) surfaces of the leaf. Parallel stomatal impressions are
mainly disposed on the abaxial side of the leaf, but a few are also seen on the adaxial side (Also Fig. 6E-F).
BAR5576a. B. A. huncoensis, profile of leaf with the margin (m) indicated. Specimen shows many rows of
stomata along the curved adaxial surface that are perpendicular to the long axis of the leaf. Discontinuous rows
of perpendicularly oriented stomata are also seen on the abaxial surface of the leaf (Also Fig. 6G). MPEF-Pb
58 10610. C. A. huncoensis, showing row of coalified, perpendicular stomata along a leaf seen in lateral view
(Also Fig. 6H). MPEF-Pb 10580. D. A. huncoensis, stomata organized in rows on abaxial face of an ovate leaf.
MPEF-Pb 10586a. E. A. pichileufensis, showing mostly parallel orientation of stomata, with some oblique.
Long axis of the leaf is up. BAR5576a. F. A. pichileufensis, stomata at higher magnification in a stacked image,
showing parallel orientation of stomata. Subsidiary cell boundaries not evident. BAR5576a. G. A. huncoensis,
showing mostly perpendicular orientation of stomata, with some oblique. MPEF-Pb 10610. H. A. huncoensis,
coalified stomata at higher magnification, oriented mostly perpendicular, subsidiary cells not well preserved.
MPEF-Pb 10580.
Figure 7. Selected fossil ovuliferous complexes of Araucaria pichileufensis (left; A, C, E) and Araucaria
huncoensis sp. nov. (right; B, .D, F). A. A pichileufensis, lectotype. Shows the triangular convex apophysis,
broad lateral wings that are rounded at the distal margins, and single embedded seed. USNM 40383e. B. A.
huncoensis, showing the narrow width and narrow, upright lateral extensions that are squared off at the distal
lateral margins. MPEF-Pb 10515. C. A. pichileufensis, syntype previously unpublished from the Berry 1938
type collection showing the characteristic triangular apophysis and broad, rounded lateral extensions. USNM
54512. D. A. huncoensis , showing the rounded apophysis and narrow lateral extensions. MPEF-Pb 10537. E. A.
pichileufensis, showing the darker central body and thin lateral extensions that are rounded on the left side that is complete, and partly missing on the right side. BAR 4352b. F. A. huncoensis, showing characteristic rounded apophysis and narrow lateral extensions. MPEF-Pb 10569.
Figure 8. Fossil ovuliferous complex (OC) details of Araucaria pichileufensis (left; A, C, E–H) and Araucaria
huncoensis sp. nov. (right; B, D, I–K).– A D: OC specimens from Río Pichileufú and Laguna del Hunco
showing ligules (lig). A. A. pichileufensis, lateral wings are incompletely preserved but the central body shows
59 triangular shape of apophysis and large triangular ligule above the seed. BAR5580. B. A. huncoensis, holotype.
Ligule is short above inverted seed. MPEF-Pb 10556. C. A. pichileufensis with oxidized preservation showing triangular apophysis and large triangular ligule above the seed (Also Fig. 8E). BAR5626. D. A. huncoensis,
showing short, broad triangular ligule above seed. MPEF-Pb 10567. E. A. pichileufensis, detail of ligule. BAR
5626. F. A. pichileufensis OC with raised striations on seed (Also Fig. 8H). BAR 5591. G. A. pichileufensis,
showing preservation of just the central body of the OC on multiple planes of the rock. BAR 5598b. H. A.
pichileufensis, detail of resin canals filled with amber casts shown under epifluorescence. BAR 5591. I. A.
huncoensis OC with partial preservation of lateral wings. Arrow indicates region where stomata were observed
(also Fig. 8J). MPEF-Pb 10533. J. Stomata on A. huncoensis OC viewed under epifluorescence, showing
perpendicular orientation. Apex is up. MPEF-Pb 10533. K. A. huncoensis, detail of seed showing striations.
MPEF-Pb 10531.
Figure 9. Fossil pollen cones of Araucaria pichileufensis (A, B) and Araucaria huncoensis sp. nov. (C, D). A.
A. pichileufensis pollen cone terminally attached to a leafy branch segment. BAR 5618b. B. A. pichileufensis
pollen cone showing helical microsporophylls and leaf-like triangular basal bracts. BAR 5619a. C. A.
huncoensis, incomplete pollen cone showing triangular subtending basal bracts. Inset shows peltate, pointed microsporophyll laminae and internal structure of cone with microsporophyll stalks (Also Fig. 10G). MPEF-Pb
10618. D. A. huncoensis, pollen cone showing pointed, rhombic laminae and long, pointed basal bracts. MPEF-
Pb 10617.
Figure 10. Fossil pollen cone details of Araucaria pichileufensis (A–F) and Araucaria huncoensis sp. nov. (G,
H). A- E: BAR 5582. A. A. pichileufensis cone showing internal part of cone body and location of pollen shown
in B, C and lanceolate basal bracts (D). B. A. pichileufensis fossil pollen visible in clumps along
60 microsporophyll stalks. BAR 5582a. C. Individual A. pichileufensis pollen grain showing nonsaccate,
inaperturate, globose-subglobose morphology. D. A. pichileufensis cuticular morphology of cone basal bract
under epifluorescence, showing mostly parallel orientation of stomata. E. A. pichileufensis counterpart of the
pollen cone specimen shown in 10A, exhibiting cone apex and peltate microsporophylls (inset). F. A. pichileufensis, detail of overlapping lanceolate basal bracts with pointed apices that emerge directly at the base of the cone. BAR 5619b. G. A. huncoensis, detail of lanceolate basal bracts for cone illustrated in Fig. 9C, with
pointed apices emerging at the base and overlapping the base of the pollen cone. MPEF-Pb 10618. H. A.
huncoensis, detail of basal bracts with broad triangular bases and showing the attachment at the cone base.
MPEF-Pb 10614.
Figure 11. Box plots showing size comparisons of fossil ovuliferous complexes (OC) between Araucaria
pichileufensis (gray; from Río Pichileufú, ~47.7 Ma) and Araucaria huncoensis sp. nov. (white; from Laguna
del Hunco, ~52.2 Ma). Median (bold line) and mean (+) shown. Box limits indicate the 25th and 75th percentiles as determined by R software; whiskers extend 1.5 times the interquartile range from the 25th and
75th percentiles. Outliers are represented by dots. Results of the two-sample t-test are given with the degrees of freedom and p-value for the difference of means. A. OC total length, in millimeters. The mean OC length of A.
pichileufensis is ca. eight millimeters larger than that of A. huncoensis. B. OC maximum width, in millimeters.
The mean OC maximum width of A. pichileufensis is ca. 13 millimeters wider than A. huncoensis. C. OC ratio
of the total length to the maximum width. Shown as raw values, but t-statistics based on log transformation to
normalize the data. D. OC position of maximum width shown as the percentage of the OC total length.
Figure 12. Size comparisons of fossil taxa Araucaria pichileufensis and Araucaria huncoensis sp. nov. with
extant Araucaria species. Measurements for extant species are from Farjon (2010) and Mill et al. (2017). A.
61 Length and width plot for mature leaves, using abaxial measurements of the adpressed/imbricate fossil leaf
forms, showing both fossil taxa overlapping each other and the smallest-leaved Eutacta species. B. OC length
and width plot showing both fossils in the smallest range of all living Sect. Eutacta. C. Pollen cone length and
width plot showing the single complete fossil specimens from each locality plotted as a star. Fossil pollen cone
specimens are on the edge of measurements in extant Sect. Eutacta, with the cone from A. huncoensis also
overlapping within the range of extant Sect. Bunya.
Figure 13. Total evidence phylogeny of Araucariaceae, with Agathis condensed. Strict consensus of the
most parsimonious tree is shown. Asterisk, fossil taxa. The extant species from New Caledonia are color-coded according to the systematic treatments of Gaudeul et al. (2012) and Escapa and Catalano (2013): “large-leaved” species = red; “small-leaved” species with an interior distribution = light blue; “small-leaved” species with a coastal distribution = dark blue. Interestingly, Araucaria huncoensis is found among the “large-leaved” clade,
although the fossils from Laguna del Hunco have some of the smallest leaves compared with any of the extant
species (Fig. 12A); however, support values are low.
Figure 1 Click here to access/download;Figure;Figure 1 final.tif Figure 2 Click here to access/download;Figure;Figure 2.tif Figure 3 Click here to access/download;Figure;Figure 3.tif Figure 4 Click here to access/download;Figure;Figure 4.tif Figure 5 Click here to access/download;Figure;Figure 5.tif Figure 6 Click here to access/download;Figure;Figure 6.tif Figure 7 Click here to access/download;Figure;Figure 7.tif Figure 8 Click here to access/download;Figure;Figure 8.tif Figure 9 Click here to access/download;Figure;Figure 9.tif Figure 10 Click here to access/download;Figure;Figure 10.tif Figure 11 Click here to access/download;Figure;Figure 11.tif Figure 12 Click here to access/download;Figure;Figure 12.tif Figure 13 Click here to access/download;Figure;Figure 13.tif Appendix S1 Click here to access/download;Online Supplemental;Appendix S1.docx
Rossetto-Harris et al. —American Journal of Botany 2019 – Appendix S1
Appendix S1. Taxa with modern systematics work, showing presence, absence, and turnover at species level between early Eocene
Laguna del Hunco and middle Eocene Río Pichileufú, southern Argentina.
Family Fossil species Laguna del Río Pichileufú Citation
Hunco
FERNS Osmundaceae Todea amissa M. Carvalho ⁎ ◌ Carvalho et al. 2013; Bomfleur and Escapa, 2019 GYMNOSPERMS Araucariaceae Agathis zamunerae Wilf ⁑⁑ ⁂ Wilf et al. 2014
Araucaria huncoensis Rossetto-Harris, sp. ⁑⁑ ◌ This study nov. Araucaria pichileufensis E.W. Berry ◌ ⁂ Berry 1938; this study
Cupressaceae Papuacedrus prechilensis (E.W. Berry) Wilf, ⁑ ⁎ Wilf et al. 2009 Little, Iglesias, Zamaloa, Gandolfo,
Cúneo et Johnson Podocarpaceae Acmopyle engelhardti (E.W. Berry) Florin ⁑ ⁎ Berry 1938; Florin 1940b; Wilf 2012 Dacrycarpus puertae Wilf ⁂ ⁎ Wilf 2012
Huncocladus laubenfelsii A. Andruchow- ⁎ ◌ Andruchow-Colombo et al. Colombo, Wilf et I. Escapa 2019 Rossetto-Harris et al. —American Journal of Botany 2019 – Appendix S1
Podocarpus andiniformis E.W. Berry ⁂ ⁎ Berry 1938; Wilf et al. 2005
Retrophyllum oxyphyllum (Freng. & Parodi) ⁂ ⁎ Wilf et al. 2017a; Wilf 2020 Wilf Ginkgoaceae Ginkgoites patagonicus (E.W. Berry) Villar ⁑ ⁎ Berry 1925, 1938; Villar de de Seoane, Cúneo, Escapa, Wilf et Seoane et al. 2015
Gandolfo Zamiaceae Austrozamia stockeyi Wilf, D. Stevenson, et ⁎ ◌ Wilf et al. 2016 Cúneo ANGIOSPERMS Ripogonaceae Ripogonum americanum R.J. Carpenter, ⁎ ◌ Carpenter et al. 2014 Wilf, Conran et Cúneo Akaniaceae Akania americana Romero et Hickey and ⁑ ◌ Romero and Hickey 1976; Akania patagonica Gandolfo, Dibbern et Gandolfo et al. 1988
Romero Asteraceae Raiguenrayun cura Barreda, Katinas, ◌ ⁎ Barreda et al. 2010, 2012 Passalia et Palazzesi Atherospermataceae Atherospermophyllum guinazui (E.W. Berry) ⁑ ⁎ Knight and Wilf 2013 C.L. Knight Casuarinaceae Gymnostoma patagonicum (Frenguelli) ⁂ ◌ Zamaloa et al. 2006 Zamaloa, G. archangelskyi Zamaloa et Rossetto-Harris et al. —American Journal of Botany 2019 – Appendix S1
Romero, and G. argentinum Zamaloa et
Gandolfo Cunoniaceae Ceratopetalum edgardoromeroi Gandolfo et ⁎ ◌ Gandolfo and Hermsen 2017 Hermsen Fagaceae Castanopsis rothwellii Wilf, Nixon, ⁑⁑ ◌ Wilf et al. 2019 Gandolfo et Cúneo and associated foliage Juglandaceae Alatonucula ignis Hermsen et Gandolfo ⁑ ◌ Hermsen and Gandolfo 2016
Menispermaceae Menispermites calderensis Jud, Gandolfo, ⁎ ◌ Jud et al. 2018 Iglesias et Wilf Monimiaceae Monimiophyllum callidentatum C.L. Knight ⁎ ◌ Knight and Wilf 2013
Myrtaceae Eucalyptus frenguelliana Gandolfo et ⁑⁑ ◌ Gandolfo et al. 2011; Hermsen Zamaloa and associated reproductive et al. 2012
material Proteaceae Orites bivascularis (Berry) Romero Dibbern ⁎ ◌ Romero et al. 1988; González et Gandolfo et al. 2007 Solanaceae Physalis infinemundi Wilf ⁎ ◌ Wilf et al. 2017b
Notes: ⁑⁑ = n >100 specimens; ⁂ = n >40; ⁑ = n >10; ⁎ = n <10; ◌ = absent; abundances are given based on amount of material cited in the literature, including Wilf et al. (2005) census data for Laguna del Hunco. Acmopyle engelhardti is morphotype TY007 in
Wilf et al. (2005). Rossetto-Harris et al. —American Journal of Botany 2019 – Appendix S1
LITERATURE CITED
Andruchow-Colombo, A., P. Wilf, and I. Escapa. 2019. A South American relative of Phyllocladus? Huncocladus laubenfelsii gen. et
sp. nov. (Podocarpaceae), from the early Eocene of Laguna del Hunco, Patagonia, Argentina. Australian Systematic Botany 32:
290–309.
Barreda, V. D., L. Palazzesi, L. Katinas, J. V. Crisci, M. C. Tellería, K. Bremer, M. G. Passalia, et al. 2012. An extinct Eocene taxon
of the daisy family (Asteraceae): evolutionary, ecological, and biogeographical implications. Annals of Botany 109: 127–134.
Barreda, V. D., L. Palazzesi, M. C. Tellería, L. Katinas, J. V. Crisci, K. Bremer, M. G. Passalia, et al. 2010. Eocene Patagonia fossils
of the daisy family. Science 329: 1621.
Berry, E. W. 1925. A Miocene flora from Patagonia. Johns Hopkins University Studies in Geology 6: 183–251.
Berry, E. W. 1938. Tertiary flora from the Río Pichileufú, Argentina. Geological Society of America Special Paper 12: 1–149.
Bomfleur, B., and I. Escapa. 2019. A silicified Todea trunk (Osmundaceae) from the Eocene of Patagonia. PalZ doi:10.1007/s12542-
019-00479-6.
Carpenter, R. J., P. Wilf, J. G. Conran, and N. R. Cúneo. 2014. A Paleogene trans-Antarctic distribution for Ripogonum
(Ripogonaceae: Liliales)? Palaeontologia Electronica 17: article 17.13.39A.
Carvalho, M. R., P. Wilf, E. J. Hermsen, M. A. Gandolfo, N. R. Cúneo, and K. R. Johnson. 2013. First record of Todea
(Osmundaceae) in South America, from the early Eocene paleorainforests of Laguna del Hunco (Patagonia, Argentina).
American Journal of Botany 100: 1831–1848. Rossetto-Harris et al. —American Journal of Botany 2019 – Appendix S1
Florin, R. 1940b. Die heutige und frühere Verbreitung der Koniferengattung Acmopyle Pilger. Svensk Botanisk Tidskrift 34: 117-140.
Gandolfo, M. A., M. C. Dibbern, and E. J. Romero. 1988. Akania patagonica n. sp. and additional material on Akania americana
Romero & Hickey (Akaniaceae), from Paleocene sediments of Patagonia. Bulletin of the Torrey Botanical Club 115: 83–88.
Gandolfo, M. A., and E. J. Hermsen. 2017. Ceratopetalum (Cunoniaceae) fruits of Australasian affinity from the early Eocene Laguna
del Hunco flora, Patagonia, Argentina. Annals of Botany 119: 507–516.
Gandolfo, M. A., E. J. Hermsen, M. C. Zamaloa, K. C. Nixon, C. C. González, P. Wilf, N. R. Cúneo, and K. R. Johnson. 2011. Oldest
known Eucalyptus macrofossils are from South America. PLoS One 6: e21084.
González, C. C., M. A. Gandolfo, M. C. Zamaloa, N. R. Cúneo, P. Wilf, and K. R. Johnson. 2007. Revision of the Proteaceae
macrofossil record from Patagonia, Argentina. Botanical Review 73: 235–266.
Hermsen, E. J., and M. A. Gandolfo. 2016. Fruits of Juglandaceae from the Eocene of South America. Systematic Botany 41: 316–
328.
Hermsen, E. J., M. A. Gandolfo, and M. C. Zamaloa. 2012. The fossil record of Eucalyptus in Patagonia. American Journal of Botany
99: 1356–1374.
Jud, N. A., A. Iglesias, P. Wilf, and M. A. Gandolfo. 2018. Fossil moonseeds from the Paleogene of West Gondwana (Patagonia,
Argentina). American Journal of Botany 105: 927–942.
Knight, C. L., and P. Wilf. 2013. Rare leaf fossils of Monimiaceae and Atherospermataceae (Laurales) from Eocene Patagonian
rainforests and their biogeographic significance. Palaeontologia Electronica 16: article 16.13.26A. Rossetto-Harris et al. —American Journal of Botany 2019 – Appendix S1
Romero, E. J., M. C. Dibbern, and M. A. Gandolfo. 1988. Revisión de Lomatia bivascularis (Berry) Frenguelli (Proteaceae) del
yacimiento de la Laguna del Hunco (Paleoceno), Pcia. del Chubut. Actas del IV Congreso Argentino de Paleontología y
Bioestratigrafía, Mendoza 3: 125-130.
Romero, E. J., and L. J. Hickey. 1976. A fossil leaf of Akaniaceae from Paleocene beds in Argentina. Bulletin of the Torrey Botanical
Club 103: 126–131.
Villar de Seoane, L., N. R. Cúneo, I. H. Escapa, P. Wilf, and M. A. Gandolfo. 2015. Ginkgoites patagonica (Berry) comb. nov. from
the Eocene of Patagonia, last ginkgoalean record in South America. International Journal of Plant Sciences 176: 346–363,
plus Erratum p. 364.
Wilf, P. 2020. Eocene “Chusquea” fossil from Patagonia is a conifer, not a bamboo. Phytokeys 139: 77-89.
Wilf, P. 2012. Rainforest conifers of Eocene Patagonia: attached cones and foliage of the extant Southeast Asian and Australasian
genus Dacrycarpus (Podocarpaceae). American Journal of Botany 99: 562–584.
Wilf, P., M. P. Donovan, N. R. Cúneo, and M. A. Gandolfo. 2017a. The fossil flip-leaves (Retrophyllum, Podocarpaceae) of southern
South America. American Journal of Botany 104: 1344–1369.
Wilf, P., M. R. Carvalho, M. A. Gandolfo, and N. R. Cúneo. 2017b. Eocene lantern fruits from Gondwanan Patagonia and the early
origins of Solanaceae. Science 355: 71–75.
Wilf, P., I. H. Escapa, N. R. Cúneo, R. M. Kooyman, K. R. Johnson, and A. Iglesias. 2014. First South American Agathis
(Araucariaceae), Eocene of Patagonia. American Journal of Botany 101: 156–179. Rossetto-Harris et al. —American Journal of Botany 2019 – Appendix S1
Wilf, P., S. A. Little, A. Iglesias, M. C. Zamaloa, M. A. Gandolfo, N. R. Cúneo, and K. R. Johnson. 2009. Papuacedrus
(Cupressaceae) in Eocene Patagonia: a new fossil link to Australasian rainforests. American Journal of Botany 96: 2031–2047.
Wilf, P., K. R. Johnson, N. R. Cúneo, M. E. Smith, B. S. Singer, and M. A. Gandolfo. 2005. Eocene plant diversity at Laguna del
Hunco and Río Pichileufú, Patagonia, Argentina. American Naturalist 165: 634–650.
Wilf, P., K. C. Nixon, M. A. Gandolfo, and N. R. Cúneo. 2019. Eocene Fagaceae from Patagonia and Gondwanan legacy in Asian
rainforests. Science 364: eaaw5139.
Wilf, P., D. W. Stevenson, and N. R. Cúneo. 2016. The last Patagonian cycad, Austrozamia stockeyi gen. et sp. nov., early Eocene of
Laguna del Hunco, Chubut, Argentina. Botany 94: 817–829.
Zamaloa, M. C., M. A. Gandolfo, C. C. González, E. J. Romero, N. R. Cúneo, and P. Wilf. 2006. Casuarinaceae from the Eocene of
Patagonia, Argentina. International Journal of Plant Sciences 167: 1279–1289.
Appendix S2 Click here to access/download;Online Supplemental;Appendix S2.docx
Rossetto-Harris et al. —American Journal of Botany 2019 – Appendix S2
Appendix S2. Living Araucaria Section Eutacta collections vouchered at the Pennsylvania State University Herbarium (PAC), made
by Gabriella Rossetto-Harris at the Montgomery Botanical Center (MBC), Coral Gables, Florida, May 23-24, 2018.
PAC Collection Species Material MBC plant Plant origins Collection accession number accession coordinates (at number number MBC; NAD 1983 UTM zone 17N) 107298 001 Araucaria luxurians Juvenile and mature 87174*C plant originally was a voucher at 2838066.5568, (Brongn. & Gris) de vegetative; female Adelaide Botanical Garden, Australia 572018.1775 Laub. cone 107297 004 Araucaria columnaris Juvenile and mature 7485*A plant grown from seed produced by 2838128.64, (G.Forst.) Hook. vegetative; male Emil Ursini in Ft. Lauderdale- first 572052.6459 cones Araucaria seed produced in Florida. 107296 002 Araucaria cunninghamii Juvenile, intermediate, 20080147*A plant originally from N.Z. Botanical 2837885.7229, Aiton ex D.Don and mature Research Institute Ltd., New Zealand 572003.6151 vegetative; female cone 107295 003 Araucaria biramulata Juvenile and mature 87177*B plant originally from New Caledonia 2838098.4808, J.Buchholz vegetative; female Locality: Thy. A. Watt 578 (S/86-13) 572029.251 cone
Appendix S3 (.tnt file)
Click here to access/download Online Supplemental (not built into PDF) Appendix S3.tnt