Permocalculus iagifuensis sp. nov.: A new Miocene gymnocodiacean alga from Papua New Guinea

Permocalculus iagifuensis, a new species of gymnocodiacean alga is described from the Miocene of the Darai Limestone Formation of Papua New Guinea. The discovery of this species greatly extends the range of gymnocodiacean algae, which previously had only been confidently recorded from the Permian and Cretaceous. It also suggests an evolutionary link to the Recent genus Galaxaura (order Nemalionales; family Chaetangiaceae), which is the only extant alga bearing a similarity to the Gymnocodiaceae. Alternatively, a closer relationship to the green udoteacean algae (e.g. Halimeda) is considered. The microfauna and other microflora associated with this new species are briefly described.


INTRODUCTION
The Darai Limestone Formation (Late Oligocene-Middle/Late Miocene) of Papua New Guinea contains abundant and diverse calcareous algae. Coralline rhodophytes are dominant, but Udoteaceae, and more rarely, Dasycladaceae, alsooccur. During the course of a review of the palaeoecological significance of calcareous algae from the lower Tf 1 largerforaminiferal biozone (cf. Adarns, 1984) (Early Miocene= biozones N6-N7 of Blow,

STRATIGRAPHY AND MICROPALAEONTOLOGY.
The Darai Limestone Formation (eg. Davies, 1983) crops out across much of the Highlands region of Papua New Guinea. The samples discussed here are from outcrops in the fold belt region south of Tari (see Fig. 1). The Darai Limestone Formation includes several bioclastic limestone types, representing a variety of environments from back-reef, through a number of reef and peri-reefal sub-environments, to fore-reef. Reefs are 1969), it was noted that gymnocodiacean algae were also present. These were referred to a new species of the genus Permoculculus which is described below.
Elliott ( 1955) erected the family Gymnocodiaceae, which presently contains only two genera, Gymnocodium and Permoculculus, for the remains of fossil plants similar to the living marine red alga Guluxuuru (order Nemaliones; family Chaetangiaceae). These fossil forms are believed to have been erect, branched plants, but are only known from perforate, calcareous segments and fragments. The fossil Gymnocodiaceae are held to be distinct from their Recent counterparts on the basis of greater calcification in the fossil forms and the discontinuous record of simi lar forms between the Permian, Cretaceous and the Recent. Gymnocodiaceae remain a poorly understood group of calcareous algae. Their peculiar discontinuous stratigraphic range, and the lack of completely analogous extant forms, has inhibited elucidation of their palaeobiology. The record of this new species, the first ever from the Neogene, is important because it points toamorecontinuous stratigraphic

150.
than previously and suggests an Location map showing the area from which Permocalculus link between fossil calcareous forms and the Recent poorly calcified Guluxuura.
was recorded.

LATE .CRETACEOUS
dominated by associations of corals and encrusting rhodophytes (e.g.Lthophyllum, Mesophyllum, Archaeolithothumnium), and occasionally bryozoans. Because of extensive rainforest cover, clear exposures of the Darai Limestone Formation are rare, thus it is difficult to visualize the architecture of the carbonate platform. This is further inhibited by extensive thrusting of the formation. In order to reconstruct the nature of the carbonate platform, isolated samples have to be examined in thin-section, and a picture of the platform gradually pieced together. Provisional analysis suggests that a broad east-west trending plat- Summary of the regional stratigraphy (not to scale). The level from whichPermocalculu.7 iagifuensis sp. nov. was recorded is highlighted.

MAASTRICH
To the east of the study area the Darai Limestone Formation may be underlain by Paleogene carbonates and sands of the Mendi broup. To the east the Darai Limestone Formation also becomes more plagic in character. Age ranges quoted for formations are variablethey may range considerably (e.g. age of top Chim Formation).
Derivation of name. Toponymic, after the lagifu Anticline, a structure in the vicinity of the sample location from where the first specimens of the new species were noted. Diagnosis. Species of Permocalculus preserved as strongly calcified segments with closely spaced cortical pores typically 0.02mm in diameter.  fig.4), often common in Barremian-Aptian platform limestones of the Middle East, is the most similar and can be considered as a homeomorph. This species is preserved as relatively large segments, up to 5mm in length and 1.75mm in width. Pores have an average diameter of O.O2mm, and adensity of 35-50 per millimeter (Elliott, 1956). P. inopinatus slightly differs from P. iagifuensis in its irregular calcification and variable pore size (0.0 1 -0.03mm reported by Johnson, 1969). The other common Cretaceous species, Permocalculus ampullacea Elliott and Permocalculus irenae Elliott (P1.2, fig.3), have small pores (typically 0.0 12mm), with a lower density per millimeter (Elliott 1958(Elliott , 1959Johnson, 1969). Of the Permian species, Permocalculus solidus (Pia) and Permocalculus digitatus Elliott are the most similar to P. iagifuensis. From the data given in Elliott (1955), these species have an average pore diameter of 0.02mm, and the segments of thallus are of a comparable size to those of P. iag$4errsis. They differ by the variable calcification and the distinctive cylindrical shape of the thallus. It appears that iterative evolution is common within the Gymnocodiaceae, since P. solidus, P . digitatus, P. inopinatus and P . iagifuensis can all be considered as homeomorphs.
It is relevant to note here that no suitable Permian or Cretaceous facies exists in the area from which P. iagifuensis is recorded, from which the species could have been reworked. Furthermore, reworking of older sediments is not common in the lower Darai Limestone Formation. Pia (1937) was the first to consider Gymnocodium, the type genus of the Gymnocodiaceae, to be related to the extant genus Galasaura of the Chaetangiaceae. Earlier, he had considered Gymnocodium to belong to the Codiaceae (=Udoteaceae) (Pia, 1920;1927). Elliott ( 1955) erected the family Gymnocodiaceae and included the genera Gymnocodium and Per~nrocalcirlus.
Explanation of Plate 2 Figs 1-2. Longitudinal sections of Hulimeda sp. x32. This species is occasionally found in association with Per-mocalcitlus iugifrrrrrsis sp. nov. Fig. 3. Permocalculus ir-enae Elliott x32. This specimen from Cenomanian sediments of the Oman Mountains is illustrated for comparison with Pcrmoculculus iagifuensis sp. nov. Fig. 4. Per-mocdc~ulus inopinutus Elliott x70. this species from the Early Cretaceous of the Oman Mountains is also illustrated for comparison with Pomocdculus ia,p$uensis sp. nov. The two species can be considered homeomorphs. Elliott maintained the view that these genera are related to Galaxaura, but included them in a separate family because of differences in the degree of calcification and the discontinuous record between the Permian and Recent forms. The view that the Gymnocodiaceae are related to Galaxaura has been upheld in subsequent papers (e.g. Elliott, 1961, Johnson, 1969. Galaxaura is very similar to Gymnocodiaceae such as Permocalculus. Species are typically formed of weakly calcified segments several millimeters in length, united to form a flexible jointed plant with a thallus up to lOcm in length. The sporangia of asexual and sexual plants are internal. Although not commonly recorded, the sporangia of Gymnocodiaceae are also internal. This criterion can be used to distinguish the family from the green Udoteaceae in which the sporangia are external. Galaxaura has a pitted, non-calcareous outer skin, beneath which is a cakareous subdermal layer. This can be seen to be penetrated by numerous pores. These have a diameter of O.Olmm (Elliott, 1955). Although not identical, this structure is very similar to that seen in species of Gymnocodiaceae.
The discovery of a new species of Permocalculus from Miocene sediments supports the contention that the Gymnocodiaceae and Galaxaura (Chaetangiaceae) are related. Gymnocodiaceae are now confidently known from Permian, Cretaceous and Miocene sediments. It may be that the Chaetangiaceae have exhibited variable calcification, and therefore preservation potential, through time, and that they have in fact experienced a continuous existence from the Permian through to the Recent. The reasons for this variable calcification are unknown, but it appears that records of Gymnocodiaceae correspond to periods of time in which extensive epeiric seas developed. If this model of variable calcification is correct, the term Gymnocodiaceae, which refers only to fossil forms, should be abandoned. Both fossil and extant taxa should be referred to the Chaetangiaceae.
Alternatively it is possible that the Gymnocodiaceae bear a closer relationship to the Udoteaceae than previously thought.
The presence of internal sporangia in the Gymnocodiaceae has been cited as a reason for dismissing any suggested relationship to the Udoteaceae. However, the presence of sporangia in the Gymnocodiaceae remains uncertain. There are few illustrations of Gymnocodiaceae where the sporangia can be clearly seen. Moreover, the two groups show similarities. Some illustrations of Gymnocodiaceae show taxa with an internal medullary and cortical arrangement like Udoteaceae (e.g. The Gymnocodiaceae remain a problematic group requir-ing further detailed studies.