Permian glossopterid scale leaves from the La Golondrina Formation ( Santa Cruz , Argentina ) : the neglected members of Glossopteris floras

The Permian fossil flora from the La Golondrina Formation (Archangelsky, 1967) from the Santa Cruz Province, Argentina, has been extensively studied since it was discovered in the early 1950s. Important contributions by different authors described a flora with abundant glossopterids, ferns, sphenophytes, and to a lesser degree, lycophytes and cordaites (Archangelsky, 1959, 1960; Archangelsky & De la Sota, 1960; Arrondo, 1972; Herbst, 1978; Archangelsky & Cúneo, 1984, 2002; Archangelsky et al., 1996a, 2004; Cariglino, 2011). Glossopterids are the dominant group in this flora, and are represented by both sterile (i.e., Glossopteris and Gangamopteris leaves, Vertebraria sp.) and fertile elements (i.e., Dictyopteridium sporiferum, Plumsteadia pedicellata, etc.) (Archangelsky, 1957, 1959, 1992; Archangelsky & Bonetti, 1963; Cariglino et al., 2009). Rev. Mus. Argentino Cienc. Nat., n.s. 15(1): 61-70, 2013 ISSN 1514-5158 (impresa) ISSN 1853-0400 (en línea)

Glossopterid scale leaves are found all over Gondwana throughout the Permian, but are more common in Upper Permian strata (Anderson & Anderson, 1985;McLoughlin et al., 2005).Despite of being a recurrent element in Glossopteris floras, their mention is usually belittled, probably because of the lack of knowledge that still surrounds these elements.Previous works by Archangelsky (1957) and Arrondo (1972) reported their presence at the La Golondrina Formation and provided brief descriptions of some of these scales.Glossopterid scales were first mentioned by Feistmantel (1881), who assigned detached scale leaves from different localities of India to the general name of Squamae.Zeiller (1902) identified a similar anastomosing venation as in Glossopteris for isolated scales, allowing him to infer relations with the glossopterids.However, even if they were frequently in association with Glossopteris, no evidence of direct connection was known, until Surange & Chandra (1974, 1975) described glossopterid reproductive organs borne on scale leaves.
The aim of this paper is to present the glossopterid scales found at the La Golondrina Formation.This work includes the revision and description of the material previously collected by Archangelsky (1957) and Arrondo (1972), the one from other collections unstudied until now, and new material recently collected during fieldtrips to the areas of Bajo de La Leona and Estancia La Juanita (Fig. 1, Table 1).Based on morphological differences, mostly on their venation patterns, three different types of scales Fig. 1.Geological maps of the Bajo de La Leona (a) and Estancia La Juanita (b) areas showing the Permian outcrops of the La Golondrina Formation fossil localities (modified from Panza, 1994 andJalfin, 1987).
are identified, and their potential function is discussed.

MATERIALS AND METHODS
The specimens from this study come from the New specimens were collected on recent field trips to the areas of Bajo de La Leona and Estancia La Juanita, where the La Golondrina Formation outcrops in the Deseado Massif, in the northeastern region of the Santa Cruz Province (Fig. 1).According to Archangelsky et al. (1996a), the formation is divided into three members, from bottom to top: Laguna Lillo, Laguna Polina and Dos Hermanos (Jalfin, 1987(Jalfin, , 1990;;Jalfin et al., 1990;Archangelsky et al., 1996a).The megafloral remains in this formation were referred to the Dizeugotheca Superbiozone (Archangelsky & Cúneo, 1984;Archangelsky et al., 1996b), and consequently a late-Early to Late Permian age was inferred (Archangelsky, 1992(Archangelsky, , 2006;;Archangelsky et al., 1996aArchangelsky et al., , 1996bArchangelsky et al., , 2004;;Andreis, 2002).The newly recovered material is housed at the Museo Provincial Regional "Padre Jesús Molina" (MPM Pb) at Río Gallegos.All the scale leaves in this study were recovered from siltstones and fine grained sandstones levels from the middle member of the unit at the localities of Laguna Polina, Laguna Castellanos, Laguna Feruglio at the Bajo de La Leona, and Laguna de los Fósiles at Estancia La Juanita (Fig. 1).Specimens were photographed using a Canon Powershot S5IS (8.0 megapixel) digital camera.For detailed analysis and illustration, specimens were photographed under strong unilateral, low-angled light with a Nikon DS.Fi1-U2 digital camera attached to a Nikon SMZ800 stereomicroscope.Scaling and adjustment of images was accomplished using Adobe Photoshop CS3 software.Detailed images were taken with a Philips XL30 scanning electron microscope.

DESCRIPTION OF THE SCALE LEAVES
Three different types of scale leaves were recognized based primarily on shape and venation pattern differences.Remarks.Type A scale MPM Pb 2609 described here is similar to the one described by Cridland (1963, p. 189, fig. 7, 36-38) for a Glossopteris flora from Antarctica.It bears resemblance in the shape and size, although the specimen from La Golondrina is slightly more elongated.The venation pattern is also similar, with dichotomizing veins tending to one side of the scale.Pant (1958, p. 159, plate 20, figs. 3, 4;text-fig 15a) described similar scales for Tanganyika (nowadays, territory of Tanzania) with a lanceolate shape, pointed apex, and a rounded (cordate) base.These scales have veins spreading from the base that arched towards the margin, occasionally forking and anastomosing.The asymmetrical shape, the tendency to one side of the scale in the venation and the presence of thick margin may be indicative of a rolled up scale to offer protection, while Pant's scales could be representing the same type of scale, unrolled.These types of scales are not as abundant as the other types, and thus, more information is needed to infer it possible function and affinities.(Thomas, 1958;Chandra & Surange, 1977;Anderson & Anderson, 1985).Within the La Golondrina Formation, a few have been previously found and described (Archangelsky, 1957;Arrondo, 1972).The size and shape of these scales is very variable, from ovate to obovate, rhomboidal to triangular, with rounded apex with a pointed tip to an acute, long apex.The base is truncated since scales are detached from the stalk on which the fertile parts are borne they generally does not get preserved (Chandra & Surange, 1977).

DISCUSSION
Up to date, several attempts to classify scale leaves have been made, various genera proposed, and diverse interpretations about their function speculated, reaching few consensuses among scholars (e.g., Pant, 1958;Chandra & Surange, 1977;Banerjee, 1984;Anderson & Anderson, 1985).Pant (1958) described scale leaves from the Upper Coal Measures (Middle Permian) of Tanganyika, but since he could not prove the presence of sporangia or seeds, he disregarded them as reproductive organs.Later, Chandra & Surange (1977) described detached scales based on shape, size, and venation pattern.They advocated that even in detached conditions, scales could be "easily assigned to their respective genera established for the reproductive organs of Glossopteris" (Chandra & Surange, 1977, p. 195).However, if the capitula or microsporangia are separated from the scale, their assignation to a certain genus could be misleading, since most fructifications are assigned based on the presence of different female or male structures, not the scale in which they are borne.Banerjee (1984) considered that scales presented significant characteristic features deserving generic status, and classified them under four genera (Gondwanolepis, Ghoshialepis, Mahudaea, Bankolaea) basing on their occurrence (e.g., singly or in group) and general physiognomy.On the other hand, Anderson & Anderson (1985) classified four different scaleleaf genera (A to D) using similar criteria as in this study (Table 2).White (1978) suggested that these deciduous scales ('squamae') together with a laminal segment composed a "scale-frond", and a group of "scale-fronds" formed a cone.Squamella is the form-genus White (1978) used to name scalefronds aggregated into cones.Cones were born at the ends of branches, and could be either sterile or fertile (White, 1978;McLoughlin et al., 2005;Adendorff, 2005).Scales would have acted as protection for the sporangia in young cones.When cones were ripe, the deciduous scales would fall and the laminal segments would elongate to allow the sporangia to hang out and disperse the spores (White, 1978, p. 477).Even if there is not enough evidence from the La Golondrina Formation that supports White's hypothesis on the organization of a cone and its development, the presence of this kind of scales may account for the existence of cone-like structures.
Here, scales are divided in three different categories principally based on their venation and basic morphologic features.This artificial separation, however, does not rule out the possibility that they all could belong to the same group (i.e., Lidgettoniaceae).
Detached scales are considered part of a glossopterid reproductive structure.Nevertheless, their function cannot be definitely specified.Most probably, scales would provide of protection to the reproductive parts it bore (e.g., microsporangia or capitula) in the early stages of reproduction.They could also act as a mean of transport (e.g.some sort of wing) away from the mother plant, once it was mature.It has also been suggested that scales protected the buds during growth of the Glossopteris plant (McLoughlin et al., 2005;Adendorff, 2005).
Up to date, the only reached consensus on glossopterid scale leaves is that they were either fertile or sterile, probably with different functions (protection of fertile structures vs. dispersal of fertile structures), and they belonged to the Glossopteris plant.It is still debated whether these structures were always grouped in cones, or displayed individually.
The apparent curl in Type A scales seem to have acted as protective, either of fertile structures or buds during the growth of the plant.Glossopteris was a deciduous plant, spending part of the year leaf-less.Maybe the growth of new leaves when climatic conditions allowed was preceded by scaly buds, as in some modern plants.Type B scale leaves, on the other hand, have been found to be part of fertile structures, bearing both micro and megasporangia in the Lidgettoniaceae (as in Eretmonia sp. and Lidgettonia sp.).Last, Type C scale leaves have been found isolated as well as aggregated into cones (White, 1978).A cone could have acted protecting the fertile structures, and once ripe, it would open and maybe act as effective seed dispersal.The origin of the circular marks on this type of scales is unknown.Some hypotheses could be that such marks are the base of trichomes or that they were produced by the contact of the scales (acting as a protective cover) of an ovule-bearing structure, being those marks the imprint of the ovules or seeds.On the other hand, it has been suggested that the texturing in Type C scales would be an artifact of preservation (Dr.Prevec, pers.comm.).In any case, until its origin is explained, it should be considered as a diagnostic feature merely for the purpose of this classification.

CONCLUSIONS
This study identifies three different types of Glossopteris scale leaves based on their venation pattern and overall physiognomy.This artificial assignation of types, however, is proposed for the sake of easiness in their identification, and by any means is intended to explain relationships or their potential functions.The different types could be representing different groups, stages on the development, functionality, or even preservational artifacts.
This artificial classification will surely change with further findings, especially of scales aggregated into cones or bearing their reproductive structures in connection with Glossopteris leaves, which will help to elucidate their true relations, as well as their real function in the Glossopteris plant.

ACKNOWLEDGMENTS
To P. R. Gutiérrez, G. Correa and E. P. Coturel for assistance in the field.To S. Archangelsky for sharing bibliography.To L. Balarino, S. Césari and an anonymous reviewer for constructive critics that improved the manuscript.To M. Iberlucea and F. Tricárico for technical assistance.To M. Fjell and B. San Juan for access to the Estancia La Golondrina and Estancia Leonardo, respectively.To H. Carrizo and E. Fernández (Fundación Miguel Lillo), L. Lezama (Museo Argentino de Ciencias Naturales), M. Tánuz Fine venation arising from the base, it diverges towards the margin, forming fine meshes.Veins arise from the disc-like, unveined base and run straight to the apex in the center, while bending and diverging towards the margin in the laterals, forming fine meshes.
Fine venation arising from the semicircular portion, occasional dichotomies and anastomoses.
Observations The name Squamella refers to the cones composed of several scales ("squamae").Some illustrations are similar to scale-leaf Gen. D of Anderson & Anderson (1985).
Others are similar to scale-leaf Type C, this work.
These scale-leaves occurred singly.
These scale-leaves occurred in sets of three.
Small bumps (in convex specimens) or scars (in concave specimens) are observed in the central portion.
Set of two scales, one concave, the other convex.(1977) for Lidgettoniaceae and Anderson & Anderson (1985) Types A and B in the venation pattern especially.
The lower and middle parts of the scales are covered by rounded, small punctuations (same as in Mahudaea?) of uncertained origin.

Fig. 3 .
Fig. 3. Comparative schematic outline and venation pattern of Types A, B and C scale leaves.All scales 5 mm.

Table 1 .
Collection material used in this study.
Chandra & Surange (1977)almost the same in all these scales.There is no midrib; a group of central veins run from the base of the scale lamina straight to the apex, and a couple of them not dividing.The lateral veins open out towards the margin, arching, forking and anastomosing in long meshes, except closer to the margin, where meshes are shorter and more numerous.The line sketches onChandra & Surange's paper (1977)show complete scale leaves that if the stalk is detached are very similar to the scales here described.All the scales represented byChandra & Surange (1977)are assigned to taxa from the Lidgettoniaceae.
Fig. 2. Type A scale leaves, a LIL Pb 99, b MPM Pb 2609.Type B scale leaves, c BAFC Pb 15540, d LIL Pb 1436, arrow points a smaller, Type C scale, e LIL Pb 101, f LIL Pb 1439.Type C scales, g MPM Pb 4454, h MPM Pb 4453, i MPM Pb 4468, j LP Pb 7121, k LP Pb 7371, l LP Pb 7371 M.E.B image.All scales 5 mm, except (l) 2 mm.central part of the scales present circular marks that are absent in the other two types.

Table 2 -
Scale leaves classifications according different authors.