Research Article |
Corresponding author: Isabel M. Van Waveren ( isabel.vanwaveren@naturalis.nl ) Academic editor: Yasen Mutafchiev
© 2019 Isabel M. Van Waveren.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
Van Waveren IM (2019) A morphometric analysis of Tobleria bicuspis, a Voltziales seed cone from the early Permian Jambi palaeoflora, Sumatra (Indonesia). PhytoKeys 119: 67-95. https://doi.org/10.3897/phytokeys.119.29555
|
Tobleria bicuspis, a coniferophyte seed cone, is described from the Jambi Palaeoflora, Sumatra of Asselian (early Permian) age. A morphometric analysis based on cones, paired fertile units, and fertile and sterile scales, demonstrates their close relationship. Small paired fertile units occur mainly in cones. Medium-sized paired fertile units occur mainly on scales. And large paired fertile units are mainly dispersed. The cones are considered female and the paired fertile units are considered to represent the seeds. The cones are composed of helicoidal, bilaterally symmetrical and deeply incised scales with paired seeds. A comparison can be made with the Voltziales female taxon Schizolepis from the Triassic and Jurassic. Tobleria is regarded as having a voltzian Voltziales affinity and dates from approximately 16 to 26 million years before any other such cones.
Asselian, Cathaysia, mesic-xeric, early conifer cones
The evolutionary trend from early to late conifers is that of a radial fertile shoot in the axil of a leaf to a fertile scale fused to, or free from, a subtending bract (
A detailed morphometric analysis of the fertile units and scales, dispersed, superimposed and inside the cones, is presented. The statistical treatment of the fertile unit, and scale width and length, is used to demonstrate the coherence of the cones and the dispersed material. Details of the fertile units, scales and cones are presented to support an interpretation of Tobleria bicuspis as a seed cone displaying features indicating a voltzian Voltziales affinity.
The Jambi palaeoflora is composed of impressions and coalified compressions that are kept at Naturalis, the Netherlands Center for Biodiversity. The samples are small in size (commonly 100 mm long, but sometimes up to 400 mm). Some cuticles could be retrieved from the samples, but no diagnostic features were observed.
The samples originated from outcrops along several rivers and their tributaries. These rivers cut into silicified Permian volcaniclastic rocks from the westerly mountainous region of the province of Jambi (
The 500 m thick Mengkarang Formation is represented by eight fining-upwards intervals, each formed at the base of the formation by a basal pyroclastic accretion wedge composed of tuffs, and overlain by tuffaceous sandstones, organic shales and limestones, while at its top the tuffaceous sandstones are more common and basalts form the base of the intervals (
Palaeomagnetic results indicate that the Mengkarang Formation, in which the Tobleria remains were found, was positioned approximately fifteen degrees north of the equator (
A total of 17 samples from the 1925 Jambi collection held Tobleria bicuspis organs. These organs consisted of cones, fertile units and scales, fertile and barren. Fertile units were defined as paired, closed, darker circular rims or paired, dark, disc-like areas, respectively. Scales were considered to be bilateral symmetrical, sub-rounded, darker areas with two apices. The possible nature of these barren scales as bracts is fully developed in the discussion. Scales were found in all 17 samples (old collection: 45308-20, 45468, 45471, 45617; and 2006 collection: KS 28). All samples from the 1925 collection had comparably fine-grained lithologies. The sample found in 2006 along the Ketidoeran Siamang stream is composed of light grey medium – to fine-grained tuffaceous sandstones and was collected from a bed of 30 to 40 cm in thickness that was formed by a volcanic mass flow. The remaining samples found in 1925 were also collected from that stream and have the same lithology. Three samples (45468, 45471 and 45311) were collected along the Karing River. These samples consist of light beige, finely banded, tuffaceous mud – and siltstone. They comprise the cones. One of the cones is under an angle to the bedding, indicating that it fell into the soft mud that kept seeds, scales and cones in the same vicinity. A number of coarser-grained tuffs from the old Naturalis Jambi collection did hold relatively large, faintly bifid scales, but preservation was too poor for any type of data acquisition.
Sample 45311 consisted of five fragments that could be reassembled into a single specimen (Aa, Ab, B, C, D). Two cones were observed in fragment 45311A, both on part (Aa) and counterpart (Ab). The cone (C1) positioned centrally is largest and its structure can best be described using the Aa side (Fig.
The three cones from sample 45311: A cone C1 from 45311 Aa, the black frame indicates the area that is detailed in Figure
Associated seeds and scales A Dispersed seeds and scales (sample 45311a) B Scale with hart shaped scar and scale with darker central area (sample 45471) C Scale seen from adaxial side showing striation pattern and scale seen from adaxial side showing seed remains (sample 45313). Scale bars: 5 mm.
Preparation of cone C1 on sample Aa allowed the visualization of its apex that was buried a bit deeper in the shale, thus indicating that the cone was positioned under at an angle to the bedding before compaction. A scale apex was also prepared on the left apical border of Cone 1 (sample Aa). Dispersed fertile scales and fertile units were found in seven samples (45308, 45310, 45311, 45313, 45320, 45468 and KS 28) (Fig.
Pictures and measurements were made with a multi-stack Zeiss SteREO Discovery V20 microscope with a Zeiss AxioCam MRc 5 for photography, and the associated program AxioVision. Measurements were preserved on the pictures and data were exported into an Excel file. The lengths and widths of the fertile units (Fig.
Diagram showing empty scale, opaque (A1) or lighter (A2), fertile units as dark discs (B1) or dark circular rims (B2), and fertile scale with lighter fertile units on darker scale (C1) or darker fertile units on lighter scale (C2) (see also Figure
For all the data handling, the program PAST (PAleontological STatistics) version 2.12 (
As the samples were too large for a fluorescence microscope, they were analyzed with the Diamondview. The instrument illuminates the surface of the sample with intense ultraviolet light specially filtered such that almost all of the light reaching the sample is of wavelengths shorter than 230 nm. To examine a sample, it is inserted into the stone holder from the port at the front of the unit. The stone is first exposed to visible light in order to focus the camera to the area of interest, after which it is illuminated with ultraviolet light and the fluorescence image is recorded to be exported to the computer (
All three samples displayed compact cones with densely packed scales. The most complete cone fragment C1 has the apical portion preserved and is positioned on the central part of sample 45311 Aa and 45311 Ab. It is 24 mm long and 8–9 mm wide at the base, representing its broadest point, and tapering apically to a width of 4–5 mm at 2 mm from the apex. The rock sample was broken at the base of the cone and the length of the intact cone remains unknown (Fig.
Fertile units and scales in cones C1 and C2, on sample 45311 Aa and Ab, were too numerous to each be rendered photographically, but were all measured and are treated (below) in the statistical analysis. On cone C1 (Fig.
A fertile scale complex occurs in facial view (Figs
The second incomplete cone C2 on sample 45311 Aa left of the central C1 cone described above, consists of six equally sized overlapping scales (Fig.
The third and last, smallest incomplete cone C3 on 45311 B, is an apical fragment where three bifid scales could be recognized (Fig.
The paired fertile units showed a variety of transitional forms from two closely adpressed hemispheres (Fig.
Details of dispersed seeds, scales and fertile scales: A paired seeds with double wall (sample 45311C) B paired seeds with triangular micropylar protrusion (sample 45311 E) C detail of triangular micropylar protrusion from paired seeds (sample 45311 E) D heart shaped paired seeds/ovules (sample 45311 D) E heart shaped paired seeds/ovules (sample 45311 B) F juxtaposed almond shaped seeds/ovules (sample 45311 B) G bicuspid scale (sample 45311 Aa) H bicuspid scale (sample 45311D) I bicuspid scale (sample 45471) J bicuspid scale with contour of two seeds/ovules (sample 45315) K bicuspid scale with darker organic contour of seed (sample 45310) L bicuspid scale with heart shaped contour of seeds/ovules (sample 45471).
The scales are bicuspid (Fig.
On the scales the pair of fertile units forms an oval consisting of two hemispheres or a heart-shape pointing adaxially on the scale (Fig.
The search for measurable organs resulted in 263 fertile units and 158 scales. Fertile units were observed in the cones, on scales or dispersed; scales were observed in the cones, dispersed fertile or empty (Suppl. materials
The set of 263 data points for the fertile unit length and width have a correlation coefficient of, respectively, 0.9976 and 0.9966 from the normal probability plot (Fig.
The histogram of the proxy for size (length x width) of all fertile units, for example, in cones, on fertile scale and dispersed (Fig.
The box charts of the fertile unit size proxies (Fig.
The distribution of the proxy for the size of the fertile units ranked in ascending order can be described with a polynomial function of the third order. In the lower part of the spectrum, fertile units are chiefly positioned in cones. In the central part, fertile units can be positioned either in cones, on dispersed scales or are free, while in the upper part of the spectrum fertile units are either free or on dispersed scales, but more commonly free (Fig.
The set of 158 data points for the scale length and width have a correlation coefficient of, respectively, 0.9975 and 0.9821 from the normal probability plot (Figs
The histogram of the proxy for size of all scales – for example, dispersed either empty or fertile – and in cones (Fig.
The box charts of the scale size proxies (Fig.
The distribution of the proxy for the size of the scales (Fig.
The scales and seeds, both dispersed and in the cone, were exposed to fluorescent light. While the sedimentary rocks holding the plant fossils displayed fluorescence, the scales and seeds were totally inert (Fig.
Absence of fluorescence in a scale from the Cone C1 from 45311 Aa: A scale under normal light showing the scale as being slightly darker than the rock B identical scale with fluorescent light showing the scale as being opaque while the rock around it shows some fluorescence (arrows indicate the scale).
The statistical analysis conducted above indicates that the paired fertile units and bifid scales belong to one population also represented in the cones. From the box charts it appears that the larger they are, the more the fertile units are dehisced. The fertile units are initially attached to the cone and then dispersed still on a scale to final dispersal without the scale. Considering size generally as a measure for maturity, it appears that more mature fertile units are more easily released. This relation is interpreted here as a natural shedding strategy and as such is applicable to an ontogenetic development from ovules to ripened seeds. The present box charts, indicating a gradual release of the fertile units, supports such an interpretation as seeds. This is often seen in younger cones; numerous Mesozoic seed cones have been found falling apart, and in Cedrus and Abies the seeds become loose as the cone matures (
Two thin walls were observed around the smaller dispersed fertile units (Fig.
Seed cones from the Carboniferous-Permian transition have either a cycadalean, a gnetalean or a peltaspermalean affinity; or, within the conifers, may belong to the Cordaitanthales, the Ferugliocladales, the Dicranophyllales or the Voltziales. Early Permian cycadalean fertile scales differ from those of T. bicuspis as their seeds are positioned laterally to the sporophyll base (
Cones with paired anatropous ovules occur in the voltzian conifers. Most typically the seeds/ovules of Tobleria bicuspis are sessile on the scale, similar to the informal group coined by
Dicranophyllales also represent early conifers with paired seeds on a helicoidally-placed scale in a compact cone, but the seeds are erect and winged (
Voltzian Voltziales are generally described with a bract, but there is little evidence available for T. bicuspis to establish the presence of a bract subtending the scale. There are more dispersed empty scales than dispersed paired seeds, thus indicating that all paired seeds are accounted for by a scale while only 55% of the dispersed paired seeds are accounted for by a second empty scale that can be hypothesized to represent a bract. Various solutions can be offered for the discrepancy between the proportion of fertile and empty scales. A taphonomic rationale can explain this bias by considering that scales are relatively lighter than seeds and that they are more easily transported, leaving relatively more seeds behind. Another solution resides in the fact that the fertile scales and bracts can be expected to be adnate or closely packed; therefore, it is difficult to distinguish one from the other in the dispersed material.
The Voltziales Schizolepis moelleri and S. planidigesita from the Middle Jurassic of the Liaoning (
Schizolepis permensis, from the Kupferschiefer of Funfkirchen, based on one scale only with a pair of inverted seeds, was described without a bract (
Comparably to Schizolepis species, Tobleria bicuspis is characterized by deeply incised scales (
Consider the suite of features shown by Tobleria bicuspis: an apically compact cone composed of helicoidally-placed scales, seeds being shed, ovules with a double wall and a small micropyle, scales with adaxially paired anatropous sessile ovules, deeply incised scales, and the absence of a seed wing. These features indicate that Tobleria bicuspis, in spite of the difficulties in unequivocally demonstrating the presence of bracts, mostly resembles the voltzian Voltziales and is reminiscent of Schizolepis, but differs from it in that the latter’s cones are generally lax (
The resemblance to Schizolepis is remarkable as it is commonly found in Mesozoic strata (
An evolutionary development from a radial ovuliferous shoot to an ovuliferous scale is a common concept for the origin of the voltzian Voltziales (
The early occurrence of Tobleria bicuspis is not as surprising as it may seem as it was found in the Mengkarang Formation where gravity flows from the extrabasinal volcanic slope environment are characterized by very early seed fern occurrences (
Voltziales with ovuliferous shoots and those with only ovuliferous scales are not necessarily mutually exclusive. Molyostrobus texanum from the early Permian of Texas has a single erect ovule carried by a flattened shoot (
The positive skewness in the histogram for seeds and scales size proxies in Tobleria bicuspis indicates that smaller fertile units and scales occur more commonly than larger ones. Fossils cones, in general, can be represented by numerous separate scales, while accidentally detached cones, on the contrary, still can hold seeds and scales (
In extant taxa, Cedrus and Abies for example, scales and seeds disperse separately when the cone ripens while in Araucaria the seed, enclosed in the cone scale, disperses as one unit (
The scale size proxy curve is also represented by three segments with three different steepnesses, but these are clearly linear: (1) small scales with a steep size distribution; (2) medium-sized scales with a moderately steep size distribution; and (3) large scales with a steep size distribution. Considering growth to be apical, the largest scales are interpreted as representing the scales near the cone base. As these large scales are chiefly empty and because a steep diagram fragment was also observed for the fertile unit size distribution, indicating that the largest paired seeds were often dehisced from the scale (Fig.
The second segment of the scale size distribution comprising empty scales, fertile scales and scales in the cones is expected to represent the central part of the cone. As the largest scales still attached to the cones are only found in the lower half of the second segment of the scale size proxy distribution, cones are expected to be twice the length of the longest fossil cone fragment found in the Jambi collection. The first segment with the smallest ones comprises most scales found in the cones and is expected to chiefly represent its apical part, last developed. Such smaller scales, when being dispersed, may also represent a constricted cone base, but there is insufficient evidence to allow for its reconstruction. In consequence, the reconstruction presented here does not comprise the very base of the cone, but a zone of large empty scales hypothesized to represent fertile scales that have shed their seeds (Fig.
In some cones of Compsostrobus, there is a fertile zone where the scales and bracts are attached at a less acute angle, a feature that is considered to be associated with maturation (
Tobleria seeds were found in association with the mesic-xeric elements from the Jambi palaeoflora. Gymnosperm leaves from these localities are Cordaites principalis, Macralethopteris hallei, Sphenopteris sp., Dicranophyllum molle and Peltaspermales. Gothanopteris boschana and taeniopterids of which the taxonomic affinities are unknown, are also part of this mesic-xeric association from the Mengkarang Formation (
Palynological analysis of only one sample from the base of the Mengkarang Formation gives good results (
The Tobleria bicuspis seeds and scales that are from the Ketidoeran Siamang River locality, representing the upper half of the Mengkarang Formation, are considered to have originated from the source area of the gravity flows and are consequently allochtonous. The cones from the Karing River, at the top of the Mengkarang Formation, have not been transported significantly and are considered to be parauchtonous. This indicates that Tobleria bicuspis grew on the slope of the Karing Volcanic Complex during what was indicated by
The Tobleria bicuspis samples were not exposed to petrographic analysis, unlike 33 samples from the Merangin section through the Mengkarang Formation. Ten out of these 33 organic-rich samples clearly contain leaf remains and indicate between 36 and 89% of inertinite, the remaining percentages being vitrinite and in four cases also low ratios of liptinite. The inertinite in these samples was chiefly composed of fusinite, but also semi-fusinite, macrinite, detrinite and inertodetrinite (
On the other hand, the seeds, scales and cones of Tobleria bicuspis displayed no fluorescence at all. Inertinite, as opposed to vitrinite and leptinite, displays no fluorescence (
Tobleria bicuspis is regarded as a compact cone, with helicoidal paired seeds on bifid scales. The cone dehisced (dispersed) and released fertile scales, scales and seeds. The most important aspect of the present description is the demonstration of the existence, in early Permian times, of a coniferophyte with scales, if not scale/bract complexes, with a derived voltzian Voltzialean architecture. The radial symmetry of ovuliferous shoots of the walchian Voltziales, common to the late Palaeozoic, cannot be detected. Tobleria bicuspis is part of the mesic-xeric Jambi palaeoflora from the West Sumatra volcanic region, where other gymnosperm taxa also appeared relatively early.
I thank Prof. Dr. JHA Van Konijnenburg for proofreading this manuscript and for her thought-provoking comments and Prof. Dr. SK Donovan, my favorite native English speaker, for his incisive review. Dr. H Zwaan is gratefully acknowledged for his help in making the fluorescence analysis possible. I also thank Martine Hermsen, Eric Jan Bosch and Bas Blankenvoort for the drawings and the diagrams. This paper was strongly improved in the light of constructive comments provided by the review of Prof. Dr. M Popa and Dr H. Anderson.
Measurements of 263 fertile units
Data type: measurement
Explanation note: Column 1: “cone” indicates the fertile units found in cones, “on scale indicates the fertile units found on dispersed scales, “free” indicates the fertile units found dispersed free from scale; column 2 gives the length of the fertile units in mm; column 3 gives the width of the fertile units in mm; column 4 gives length x width and represents a proxy for the fertile unit size.
Measurements of 158 scales
Data type: measurement
Explanation note: Measurements of 158 scales: Column 1: “In cone” indicates the scales found in cones, “fertile dispersed” indicates the dispersed scales that are fertile, “empty dispersed” indicates the empty dispersed scales. Column 2 gives the length of the scales in mm; column 3 gives the width of the scales in mm; column 4 gives length x width and represents a proxy for the scale size.