1. Introduction
Anacardiaceae and Burseraceae (Sapindales) provide an excellent opportunity for investigating the biogeographic history of tropical diversification and the relative importance of movement and climatic adaptation in angiosperm evolution (Weeks et al., 2014). These families occur on every continent except Oceania and Antarctica and are major elements of temperate, seasonally dry tropical forests and tropical wet forests (Gentry, 1988; Pennington et al., 2010). Although these two families have approximately the same number of species, the Anacardiaceae occupy a wider range of habitats (Weeks et al., 2014).
Anacardiaceae/Burseraceae have a long evolutionary history. Xie et al. (2014) suggested the divergence occurred approximately 73 Ma, based on the oldest fossil record of Anacardiaceae (Estrada-Ruiz et al., 2010). The fossil record of Anacardiaceae is extensive worldwide (Stevens, 2001 onwards), with approximately 80 specimens associated with the family (e.g., Kruse, 1954; Wheeler and Manchester, 2002; Martínez-Cabrera and Cevallos-Ferriz, 2004; Gregory et al., 2009; Estrada-Ruiz et al., 2010; Pérez-Lara et al., 2017; Woodcock et al., 2017). The majority of these records are from South America and Asia (Weeks et al., 2014).
There are only two reports of fossil Anacardiaceae from Panama, both discovered in the Azuero Peninsula: Dracontomelon L. endocarps from the Eocene Tonosí Formation (Herrera et al., 2012) and one fossil wood that resembles Anacardiaceae/Burseraceae from the Oligocene-Miocene Santiago Formation, found in the surrounds of the Ocú town (Jud and Dunham, 2017).
Here we report the largest fossil trunk discovered in Panama and probably, Central America, to date. The wood has several features that support an affinity to the Anacardiaceae, although they are not found in any specific extant genus. Consequently, we erect a new fossil-genus and species of Anacardiaceae.
This new record is presented as additional evidence that Central America (including Mexico) was a divergence center of the Anacardiaceae.
2. Materials and methods
The two specimens reported herein were collected in the Veraguas province in Panama. STRI 44038B is a large trunk collected on a private farm in Llano de La Cruz (latitude 08° 09’ 4.7” N; longitude 80° 53’ 11.2” W) (Figure 1).
The preserved holotype fossil trunk is ~ 20 m in length and 80 cm in preserved diameter (Figure 2A to 2C). The paratype STRI 45789, was collected in Boquerones, San Francisco (latitude 08° 13’ 46.4” N; longitude 80° 51’ 45.1” W). The paratype has a size of approximately 9.3 cm in wide and 8.5 cm in length.
2.1 Geological Context
STRI 44038 is the largest fossilized trunk found in Panama to date. The trunk is lying within layers of sandy mudrock and sedimentary breccia, parallel to bedding. Fining upward sequences indicate fluvial environments as the main depositional system. STRI 45789, as all the others fossil woods we have collected from this area, is found as a float specimen on cattle farms covered with vegetation, as the most resistant ‘clasts’ from differential weathering. This area has been mapped as part of the Oligocene to Miocene Santiago Formation, sometimes referred to as Macaracas Formation that outcrops in the Macaracas Basin (Buchs et al., 2011). Kolarsky et al. (1995) reported samples of pollen, foraminifera, and nannofossils from the Macaracas Basin in central Azuero and concluded that they support a late Oligocene to early Miocene age, although the preservation of the specimens was poor.
The age of the Santiago Formation has not been clearly determined due to its stratigraphic complexity, lack of good outcrops, and absence of radiometric elements for dating. We are currently testing detrital zircons and studying the biostratigraphy.
2.2 Accession data, specimen preparation, and imaging
Petrographic thin sections of fossil material were prepared in transverse (TS), radial longitudinal (RLS) and tangential longitudinal (TLS) sections. Sections were mounted on glass slides using EpoFix resin, ground to a thickness of ~30 μm, and coverslips were affixed with Canada balsam. Material was observed and imaged using an Olympus BX53 and digital camera SC100 with sensor CMOS of 10.5 Mpix and a Zeiss AXIO Zoom V16, photographed with an AxioCam MRc5 camera.
The fossil woods were compared with the available images of modern and fossil woods in the Inside Wood Database (IWD; http://insidewood.lib.ncsu.edu; Wheeler, 2011) and literature (e.g., Terrazas, 1994, 1999; León, 2003, 2014; Woodcock et al., 2017).
2.3. Iawa feature description and coding
We described the fossil wood specimens following the International Association of Wood Anatomists (IAWA) List of Features for Hardwood Identification (IAWA Committee, 1989). For quantitative data of vessel frequency, ray density, and vessel grouping, we made measurements in 10 different fields of 1 mm2 of area. For other quantitative features, we obtained a minimum of 25 measurements (mean vessel diameter, intervessel pit diameter, vessel-ray parenchyma pit diameter, vessel element length, and ray height). In the descriptions, we give a list of the IAWA character code numbers. We use the symbol ‘?’ to indicate that there is uncertainty as to whether the feature is present or absent and “v” to indicate that the feature is variable in occurrence.
3. Results
3.1. Systematic Paleobotany
Order - Sapindales Dumortier
Family - Anacardiaceae Lindley
Genus - Llanodelacruzoxylon Rodríguez-Reyes, Estrada-Ruiz et Gasson, gen. nov.
Species - Llanodelacruzoxylon sandovalii Rodríguez-Reyes, Estrada-Ruiz et Gasson, sp. nov. (Figures 2 to 4).
Diagnosis: Growth rings are indistinct; wood is diffuse porous; vessels are solitary combined with a few short radial multiples; perforation plates simple; intervessel pitting alternate; polygonal and medium to large; vessel-ray parenchyma pits mainly circular with reduced borders, axial parenchyma apotracheal diffuse, scanty paratracheal to vasicentric, and slightly aliform, rays heterocellular, mostly uniseriate and very rarely biseriate, non septate fibers, thin to thick walled, large solitary rhomboidal crystals very abundant in body and upright ray cells.
Species diagnosis: As for the genus
Etymology: The generic name refers to the Llano de la Cruz locality, where the holotype was collected. The specific epithet is in recognition of Mr. Carlos Sandoval, who provided samples of the large trunk for this study.
Holotype: STRI 44038 B
Paratype: STRI 45789
Repository: Center for Tropical Paleoecology and Archaeology, Smithsonian Tropical Research Institute, Panama City, Panama.
Type locality: Llano de la Cruz, Veraguas (Azuero Peninsula).
Latitude 08° 09’ 4.7” N and longitude 80° 53’ 11.2” W.
Description in IAWA feature numbers: 2p, 5p, 10a 11a, 13p, 22p, 31p, 40a, 41a, 66p, 78p, 79p, 98a, 99a, 100a, 130a, 136p, 189p.
Description: Description based on two samples. Growth rings indistinct to absent. Wood is diffuse porous; vessels are solitary (66%) combined with a few short radial multiples (34%) of 2-3 (-5) (Figures 3A to 3D), circular to oval in outline (Figure 3D); vessel mean tangential diameter 154 (range = 96 - 205, SD = 35.8) μm (Figure 3D); 6 (range = 4 - 9, SD = 1.5) vessels per square millimeter; perforation plates are simple (Figure 3E); intervessel pits are alternate and crowded, polygonal and medium to large (mean pit diameters 8 - 12 μm) (Figure 3F); vessel-ray parenchyma pits with reduced borders, round in shape (mean pit diameters 7.5 - 12 μm) (Figure 3H); mean vessel element length is 365 (range= 203 - 504, SD =87)μm. Axial parenchyma is apotracheal diffuse and scanty paratracheal to vasicentric (Figures 3B to 3D). Some vessels possess slightly aliform parenchyma (Figure 3C). Parenchyma strands are 6-celled (Figure 3G).
Rays heterocellular, mostly uniseriate, very rarely biseriate (Figure 4A and 4B). Mean ray frequency is 19 per mm (range = 15 - 23, SD = 2.7) (Figure 4A and 4B), composed of mixed (procumbent and upright) cells throughout the ray body (Figure 4C and 4D). Non-septate fibers (Figure 4B), thin to thick walled. Solitary, non-chambered, rhomboidal crystals very abundant in body and upright ray cells in radial alignment in procumbent ray cells (Figure 4C). The crystals are large, 43 to 50 μm in length and 22 to 39 µm in width (Figures 4C to 4D).
4. Discussion
4.1 Inside wood search
Burseraceae and Anacardiaceae are anatomically very similar, however Anacardiaceae can be distinguished by having more abundant axial parenchyma, more frequent septate fibers, and fewer occurrences of radial canals (Terrazas, 1994; Bell et al., 2010).
The most restrictive search we obtained in the IWD included the following characters: growth rings indistinct or absent (2p); wood diffuse porous (5p); vessels in radial multiples of 4 or more common absent (10a); vessel clusters common absent (11a); perforation plates simple (13p); intervessel pits alternate (22p); vessel-ray pits with much reduced borders (31p); vessel mean tangential diameter <50 µm absent (40a); vessel mean tangential diameter 50 -100 µm absent (41a); fibers non-septate (66p); axial paratracheal parenchyma scanty (78p), Axial paratracheal parenchyma vasicentric (79p); rays 4-10 cells wide absent (98a); rays >10-seriate absent (99a); Rays with multiseriate portion(s) as wide as uniseriate portions absent (100a); radial canals absent (130a); prismatic crystals present (136p); tree (189p). This com-bination of traits led to 18 results that included mostly species of Anacardiaceae and Myrtaceae. Myrtaceae can be ruled out because of exclusively solitary vessels, greater abundance of parenchyma compared to this fossil wood and vessel ray pitting with distinct borders. The other results consisted of only one or two species in families such as: Erythroxylaceae, Gentianaceae, Juglandaceae, Malvaceae, Moringaceae, Oxalidaceae, Phyllanthaceae and Rhamnaceae. Abundance of banded apotracheal parenchyma rules out the Erythroxylaceae and Malvaceae. The single result of Rhamnaceae is ruled out by its occasional distinct growth ring boundaries, vessel grouping common and more abundant parenchyma. Storied rays distinguish Moringaceae. Sarcotheca glauca (Oxalidaceae) possesses intervessel pitting large (>10µm) and are mainly shrubs, and not trees as the fossil. different from the fossil. Bridelia micrantha (Phyllan-thaceae) has several features distinct from the fossil, occasional reticulate perforation plates, vessels in diagonal pattern and vessel grouping common, vessel-ray pits of different sizes in the same ray cell and septate fibres present.
After the comparison we conducted, Anacardiaceae is the family that most resembles the fossil studied herein. We detail the genus identification further in the discussion.
4.2 Comparison with Anacardiaceae Genera
The mosaic of traits described from these two specimens, growth rings indistinct, wood diffuse porous, vessels solitary combined with very few short radial multiples, perforation plates simple, intervessel pitting alternate, vessel-ray parenchyma pits mainly circular with reduced borders, axial parenchyma apotracheal diffuse and scanty paratracheal to vasicentric, rays heterocellular, mostly uniseriate, very rarely biseriate rays, non septate fibers and abundance of large solitary rhomboidal crystals in procumbent and upright ray cells lead to the Anacardiaceae.
The most restrictive search in IWD included the following Anacardiaceae taxa: Faguetia falcata, Mangifera caesia, Mangifera cf. griffithii and Protorhus sp. Faguetia falcata has vessel-ray pitting in palisade and gash like, tyloses common and occasional bands of parenchyma. All these characters are absent in the fossils. The most noticeable differences between the Mangifera species from the results and LI. sandovalii, are based on wider rays and bands of parenchyma and occasionally it has a low ray density (<4 per linear mm). Protorhus possesses scalariform perforation plates and septate and non- septate fibers, whereas the fossil has simple perforation plates and non-septate fibers.
At tribe level, this fossil shares several characters with the Anacardieae (sensu Engler), such as growth rings indistinct, wood diffuse porous, vessels with no arrangement, intervascular pitting polygonal and large (>10 µm), mean vessel density (<5 per mm2), non septate fibres, paratracheal parenchyma present, but no abundant; rays uniseriate to biseriate and rays heterocellular. We stress on the intervessel pitting size that is larger in the Anacardieae compared to the other tribes (Terrazas, 1994)
From the results in Inside Wood, we note that none of the taxa match the fossil. We then consulted the largest Anacardiaceae micromorphology slide collection from Instituto de Biología, Universidad Nacional Autónoma de México, under supervision of Dra. Teresa Terrazas. From that revision it was revealed that although all characters in this fossil occur in Anacardiaceae, no modern genus possesses the diagnostic features. Moreover, the pattern of vessel-ray pitting in the fossil is not common in modern Anacardiaceae. Although vessel-ray pits are all simple, circular and small (7.6 to 12.0 µm), most anacards have larger pits in palisade and gash-like pattern. We mention Mangifera as one of the genera that has a similar pattern of vessel-ray pits. Mangifera also has a good number of species with uniseriate rays and abundant crystals in procumbent and upright ray cells (Dra. Teresa Terrazas com. pers.). There are some genera that have pits smaller than 8 µm, for example Cotinus, Dobinea, Haplorhus, Lithraea, Rhus, Schinus, Toxicodendron, and Trichoscypha. These genera do not have aliform parenchyma. Trichoscypha has unilateral winged-aliform, but it presents few uniseriate rays as well as radial canals.
4.3 Comparison with other fossil woods of Anacardiaceae
There are abundant records of anacardiaceous fossil woods. More than eighty wood types have been reported worldwide (Franco and Brea, 2008; Franco, 2009; Pujana, 2009; Gregory et al., 2009; Cheng et al., 2012; Mendez-Cardenas et al., 2014; Shukla and Mehrotra, 2016; Allen, 2017; Perez-Lara et al., 2017; Wheeler et al., 2017; Woodcock et al., 2017). A good number of the records come from Europe, Asia and South America (Ramírez et al., 2000).
We surveyed the Anacardiaceous fossil woods using the IWD, in order to elucidate whether LI. sandovalii resembles other previously reported fossil taxa. We summarize this survey in Table 1. We did not find any fossil wood that has all the characters in these specimens.
Taxon | Distinct characteristics | Age | References |
---|---|---|---|
Anacardioxylon | Septate fibres, rays 4-10 seriate, radial canals | Cretaceous-Miocene | Inside Wood, 2019 |
Anacardium | Septate fibres, vascular/vasicentric tracheids | Eocene | Woodcock et al., 2017 |
Astroniumxylon | Radial canals, septate fibres | Pliocene | Franco and Brea, 2008; Franco 2009 |
Bosquesoxylon | Septate fibres, rays 4-10 seriate | Eocene | Pérez-Lara et al., 2017 |
Bouea | Banded parenchyma, vessels 50-100 µm | Miocene | Agarwal, 1988 |
Buchanioxylon | Septate fibres, radial canals | Miocene | Roy and Ghosh, 1979-80 |
cf Mangifera | Aliform parenchyma, marginal parenchyma | Eocene | Woodcock et al., 2017 |
Dracontomeloxylon | Septate fibres, rays 4-10 seriate | Cretaceous- Miocene | Inside Wood, 2019 |
Dracontomelumoxylon | Tyloses common, septate fibres, rays 4-10 seriate | Cretaceous | Inside Wood, 2019 |
Edenoxylon | Wood semi-ring porous, radial canals | Paleocene-Eocene | Allen, 2017 |
Glutoxylon | Banded parenchyma, radial canals | Miocene-Pliocene | Inside Wood, 2019 |
Holigarnoxylon | Tyloses common, aliform parenchyma | Eocene-? Miocene | Inside Wood, 2019 |
Lanneoxylon | Septate fibres, rays 4-10 seriate, radial canals | Miocene | Inside Wood, 2019 |
Loxopterygium | Tyloses common, septate fibres, radial canals | Oligocene | Méndez-Cárdenas et al., 2014 |
Mangiferoxylon | Growth rings distinct, aliform parenchyma, banded parenchyma | Miocene-Pliocene | Inside Wood, 2019 |
Maureroxylon | Growth rings distinct, septate fibres | Eocene | Wheeler and Manchester, 2002 |
Melanorrhoeoxylon | Tyloses common, banded parenchyma, homocellular rays, radial canals | Miocene | Inside Wood, 2019 |
Pistacioxylon | Growth rings distinct, Vessels in diagonal pattern, helical thickenings, tyloses common, radial canals | Miocene | Cheng et al., 2012 |
Post Oregon site | Growth rings distinct, wood ring-porous, vessels in tangential bands | Eocene | Inside Wood, 2019 |
Resinaxylon | Vessels in diagonal pattern, septate fibres, radial canals | Oligocene | Pujana, 2009 |
Rhus | Wood semi-ring porous, vessel clusters common, septate fibres, | Eocene-Oligocene | Inside Wood, 2019 |
Schinosipxylon | Tyloses common, rays 4-10 seriate, septate fibres, radial canals | Pliocene | Inside Wood, 2019 |
Sclerocaryoxylon | Septate fibres, axial parenchyma absent, radial canals | Miocene | Biondi, 1981 |
Swintonioxylon | Tyloses common, banded parenchyma, radial canals | Miocene-Pliocene | Inside Wood, 2019 |
Tapirira | Tyloses common, septate fibres, radial canals | Eocene | Inside Wood, 2019 |
Terrazoxylon | Septate fibres, radial canals | Eocene | Wheeler and Manchester, 2002 |
Comparative | fossil genera of Burseraceae | ||
Boswellioxylon indicum | Vessel-ray pitting similar to intervessel pits, radial canals | Cretaceous | Dayal, 1965 |
Burseroxylon 06010441703 | Septate fibres, rays 4-to-10 seriate | Miocene to Pliocene | Inside Wood, 2019 |
Canarioxylon | Septate fibres, tyloses common | Miocene to Pliocene | Inside Wood, 2019 |
Canarium | Radial canals, septate fibers, ays 4-to-10 seriate | Miocene to Pliocene | Awasthi and Srivasta, 1990 |
Sumatroxylon | Septate fibers, radial canals, growth rings distinct | Miocene | Den Berger, 1923 |
Tetragastroxylon | Tyloses common, septate fibers radial canals | Miocene | Martinez-Cabrera et al., 2006 |
We remark that the vessel-ray pitting pattern in our fossil wood is the same as observed in Anacardium incahuasi from early Eocene of the Fossil Forest Piedra Chamana in Peru (Woodcock et al., 2017). However, A. incahuasi differs from our wood in having septate fibers, vascular vasicentric tracheids, rays 1-3 cells wide. Therefore, we erect a new fossil anacard genus and species.
4.4. Paleoecological and biogeographical remarks
STRI 44038B preserved trunk is ~ 20 m in length and 2.5 m wide. The dimensions suggest this “big tree” was probably ~35 m high (Figures 2A to 2C), according to Niklas’ (1995) equation. We confirmed the sandstones where it was lying were described as part of the Santiago Formation because of their properties and geological observations in the field.
Anacardiaceae and Burseraceae are good families to understand migration patterns of angiosperm families (Angiosperm Phylogeny Group, 2016). Both families probably originated in Asia and diverged during the Cretaceous (Xie et al., 2014). Long distance dispersal played a key role in migration to North and South America (Weeks et al., 2014; Xie et al., 2014). Anacardiaceae displays a much wider diversity of fruit morphology and tolerance to a more diverse range of habitats than Burseraceae. This difference may help explain why Anacardiaceae has become more widespread and has successfully occupied a wider range of biomes than Burseraceae (Weeks et al., 2014).
Weeks et al. (2014) performed ancestral area reconstructions highly congruent with the fossil record. Anacardiaceae diversified during the Cretaceous and expanded into sub-Saharan Africa. From there, the route to conquer South America probably took place during the Paleogene. It is suggested that Anacardiaceae continued to steadily colonize Eurasia and temperate zones during the Miocene, when it had a widespread geographic range and likely a diversity of climatic tolerances (Weeks et al., 2014). The fossil record of Anacardiaceae was augmented especially during the Oligocene in Central America. Reports of Anacardiaceae decrease towards the Miocene.
Herrera et al. (2012) found endocarp rich assemblages in the Eocene Tonosí Formation (Azuero, Panama) that suggested a diverse rainforest. From this Formation, they identified only one Anacardiaceae (Dracontomelon) from permineralized endocarps, that is a genus only inhabiting Asia and Africa today. Based on the absence of this Anacardiaceae genus in fossil beds from the lower Miocene Cucaracha Formation, they conclude this could represent an example of local extinctions and that long time dispersal events between New and Old world forests were common in the Paleogene. Our report of LI. sandovalii from Panama, Central America adds to the understanding of the historical biogeography of Anacardiaceae and helps support the theory that Central America (including Mexico) was a divergence center of the family in the past.