Observations on the fossil resting spore morphogenus Peripteropsis gen. nov. of the marine diatom genus Chaetoceros (Bacillariophyceae) in the Norwegian Sea

I. Suto. 2005. Observations on the fossil resting spore morphogenus Peripteropsis gen. nov. of the marine diatom genus Chaetoceros (Bacillariophyceae) in the Norwegian Sea. Phycologia 44: 294–304. The morphology and taxonomy of the fossil diatom resting spore morphogenus Peripteropsis gen. nov. from the lower Oligocene through middle Miocene sediments of Deep Sea Drilling Project Site 338 in the Norwegian Sea were examined. The new genus Peripteropsis is characterized by elongated processes on the shoulder or centre of its valves and contains four species including one new species and three new combinations: P. tetracladia sp. nov., P. trinodis comb. nov., P. norwegica comb. nov. and P. tetracornusa comb. nov. Peripteropsis tetracladia, the oldest species of the genus, arose in the early Oligocene in the stratigraphic records of the Norwegian Sea, and all Peripteropsis species including the last species, P. tetracornusa, became extinct by the earliest late Miocene. Some species are biostratigraphically useful in the Norwegian Sea and the North Pacific. Moreover, two similar species which may belong to Peripteropsis, ‘Periptera’ schraderi and ‘Periptera’ petiolata, are also described.


INTRODUCTION
Chaetoceros Ehrenberg 1844 is one of the largest and most diverse of the marine planktonic diatom genera ( Van-Landingham 1968;Rines & Hargraves 1988;Hasle & Syvertsen 1996). It plays an important role in marine primary production in nearshore upwelling regions and other coastal areas. Most species of the subgenus Hyalochaete Gran are known to form resting spores under various unfavourable conditions, such as nutrient depletion, darkness, and low temperature (e.g. Durbin 1978;Garrison 1981;Hargraves & French 1983;Kuwata & Takahashi 1990;Kuwata et al. 1993;Oku & Kamatani 1995, 1997, 1999McQuoid & Hobson 1996). On the other hand, resting spores of the other subgenus Phaeoceros Gran are reported for only one species, Chaetoceros eibenii (Grunow) Meunier. The resting spores of Chaetoceros are differentiated from the weakly silicified vegetative frustules by possessing heavily silicified valves, and they occur frequently in nearshore sediments in association with other fossil diatom valves.
In previous studies, many fossil resting spore genera such as Dicladia Ehrenberg 1854, Periptera Ehrenberg 1854, Syndendrium Ehrenberg 1854, Xanthiopyxis Ehrenberg 1854, Liradiscus Greville 1865 and Monocladia Suto 2003a have been described, and diatomists have realized that they may belong to the genus Chaetoceros. However, except for studies such as Gersonde (1980), Akiba (1986), Lee (1993) and Suto (2003aSuto ( , b, 2004, the investigation of the taxonomy and biostratigraphy of fossil resting spores has been limited, because their weakly silicified vegetative valves are not usually pre-served as fossils in sediments and the classification of resting spores was thought to be difficult or impossible. Since Periptera was erected by Ehrenberg (1844Ehrenberg ( , 1854, several species have been described (e.g. Abbott & Andrews 1979;Dzinoridze et al. 1979). Recently, Suto (2003b) described Periptera tetracornusa from upper lower Miocene to Pleistocene sediments in the North Pacific Ocean and demonstrated its biostratigrapic utility. However, the name 'Periptera' is invalid because the angiosperm Periptera DC 1824 has priority over the diatom resting spore morphogenus 'Periptera'. Therefore, the invalid fossil diatom resting spore morphogenus name 'Periptera' is renamed as Peripteropsis in this study.
Four Peripteropsis species, including one new species and three new combinations transferred to new genus from 'Periptera', are described herein from lower Oligocene through middle Miocene sediments at the Deep Sea Drilling Project (DSDP) Leg 38, Site 338 in the Norwegian Sea (Figs 1, 2) to clarify the systematic taxonomy of this genus, which might appear in the early Oligocene and expanding to the Atlantic and North Pacific.

MATERIAL AND METHODS
In this study, 80 samples of a section of middle Eocene through middle Miocene sediments from DSDP Leg 38, Site 338 (67Њ47.11ЈN, 05Њ23.26ЈE; water depth 400.8 m) were examined (Fig. 1). The diatom biostratigraphy of this site was reported in detail by Schrader & Fenner (1976) and Dzinoridze et al. (1978). Most samples investigated in this study include well preserved and abundant resting spore assemblages. The preparation and counting methods of resting spores used here follow Koizumi & Tanimura (1985), Akiba (1986) and Suto (2003b).

RESULTS
Peripteropsis species were moderately well preserved in the studied material. The cell/valve counts and stratigraphic distribution of each species are shown in Figs 3, 4 and Table 1. The stratigraphic ranges and ages are described relative to the Neogene north Pacific Diatom Zone (NPD) code of Akiba (1986) and Yanagisawa & Akiba (1998) in the Miocene and the diatom zones of Schrader & Fenner (1976) in the Oligocene and Eocene. The morphogenus name Peripteropsis is used for the fossil resting spores according to Articles 3.3 and 3.4 of the International Code of Botanical Nomenclature (Greuter et al. 2000).

Genus Peripteropsis gen. nov.
GENERIC TYPE: Peripteropsis tetracladia sp. nov. DESCRIPTION: Frustule heterovalvate. In girdle view epivalve hyaline and flat to slightly convex, mantle distinct. Epivalve face possesses some elongated processes at the shoulder or the centre of the valve. Mantle of epivalve hyaline and smooth. Hypovalve, hyaline, convex. Mantle of hypovalve with a single ring of puncta at its base. In valve view, valve outline broadly elliptical to subcircular.
DESCRIPTION: Valve ovate in valve view, apical axis 37-52 m, pervalvar axis 9-11 m. Epivalve hyaline, convex at the centre, with three slender post-like scarcely dichotomous branching processes, with valve mantle. The slender post-like scarcely dichotomous processes hyaline, located at the conical centre and at each apex of the epivalve. Cross-sectional shape of slender post-like scarcely dichotomous processes circular. Valve mantle hyaline. ETYMOLOGY: The Latin trinodis means 'three knots'. REMARKS: Hanna (1927) placed this species in the genus Dicladia, but the species should not be included in Dicladia because it does not possess two conical elevations with dichotomous branching processes at its tips on the epivalve (see Suto 2003a). This species is characterized by the three slender post-like processes and its flat epivalve and belongs to the genus Peripteropsis. Frustule not observed by Hanna (1927) or Barron & Mahood (1993). (c) tricornate process; (d) slender post-like scarcely dichotomous branching process; (e) tricornate dichotomous branching process; (f) spiny process; (g) thin and wide dichotomous branching process; (h) slender central process; (i) prominent central process; (j) mantle; (k) a single ring of puncta; (l) epivalve; (m) hypovalve. All sketches were made using LM except for Peripteropsis trinodis, which is drawn after illustrations in Barron & Mahood (1993), P. jouseae after Dzinoridze et al. (1979, fig. 164), and 'Periptera' petiolata after LM illustrations in Abbott & Andrews (1979). Sheshukova-Poretzkaya (1949, p. 207, pl. 98, fig. 11a nec fig. 11b DESCRIPTION: Frustule heterovalvate, apical axis 16.0-26.5 m, transapical axis 6.5-13.0 m, pervalvar axis 5.5-9.5 m. Valve narrowly to broadly elliptical in valve view. Epivalve hyaline, slightly convex in the centre, with four, sometimes several, numbers (see Figs 13,14), tricornate processes, and with valve mantle. Tricornate processes hyaline, with dichotomous branching processes at their tips, curved near their apices. Mantle of epivalve hyaline, high. Hypovalve hyaline, flat, with two strong spines at the apices. Mantle of hypovalve hyaline, high, with a single ring of puncta along the edge of the mantle. SIMILAR TAXA: This species is differentiated from P. tetracornusa and P. trinodis by its processes with triangular cross-sectional and dichotomous branching processes at its tips. This species is very similar to P. norwegica because it possesses dichotomous branching processes at the edge of the valve, but it is identified by the four tricornate processes on its epivalve. This species differs from 'Periptera' schraderi and 'P.' petiolata by lacking central processes.

Peripteropsis tetracladia sp. nov.
STRATIGRAPHIC OCCURRENCE: This species is recognized only at DSDP Site 338 and occurs rarely but continuously from the early Oligocene to the middle Miocene.
DESCRIPTION: Frustule heterovalvate, apical axis 40-50 m, pervalvar axis 5.0-9.0 m. Valve oval in valve view. Epivalve hyaline, slightly convex in the centre, with two tricornate processes at valve shoulder, and two slender central processes, with valve mantle. The tricornate processes hyaline, and curved near their apices. Epivalve mantle hyaline and high. Hypovalve hyaline, flat, with a single ring of puncta at its base.  et al. (1978, 1979) this species was noted in the upper Oligocene to the lower Miocene in the Norwegian Sea, but in this study it was not recognized in that interval.
REMARKS: Type specimen of Periptera schraderi Jousé in Dzinoridze et al. (1979, fig. 163, holotype) is identified to Peripteropsis tetracladia possessing broken tricornate processes; therefore Periptera schraderi is invalid. Therefore, Periptera schraderi Jousé in Dzinoridze et al. (1979, fig . 164), the other type figure of P. schraderi, must be renamed because it possesses two slender central processes. However, it is not determined whether or not this species belongs to Peripteropsis in this study.
In this study, this type species has not occurred, but very similar species are recognized ('Periptera' schraderi?,Figs 41,42,49). The specimen in Figs 41, 42 possesses one central slender process, and one in Fig. 49 lacks central slender process, but each specimen possesses two tricornate processes on the valve shoulder and may belong to 'Periptera' schraderi.  1968, p. 138, pl. 38, fig. 14; Periptera petiolata Andrews in Abbott & Andrews 1979, p. 248, pl. 4, figs 30-34, pl. 8, fig. 1; Periptera sp. of Lee 1993, p. 43, pl. 1, fig. 14, pl. 3, fig. 7. DESCRIPTION: Frustule heterovalvate, apical axis 45-70 m, transapical axis 12.0-18.5 m, pervalvar axis 15-30 m. Valve narrowly elliptical to elongate rectangular in valve view. Epivalve hyaline, slightly convex, sometimes slightly concave in central area, with distinct mantle, which is hyaline and high. Hypovalve hyaline, with a central prominent process 8-10 m in height, with a sheath surrounding the hypovalve margin. A prominent process flange is present (Fig. 47) around the margin and its top is flat at the centre. Mantle of hypovalve hyaline, high, with a single ring of puncta at its base.  1 Numbers indicate individuals encountered during counts of 100 resting spore valves; ϩ indicates epivalves encountered after the count; blank indicates absence of any taxa. Diatom zones and NPD codes in the Miocene after Yanagisawa & Akiba (1998), and diatom zones in the Oligocene and the middle Eocene after Schrader & Fenner (1976). Preservation: G, good. Abundance: A, abundant; C, common; R, rare. T, total number of resting spore valves counted. REMARKS: It is not clear whether or not this species belongs to the genus Peripteropsis because it lacks elongated processes on the epivalve. The prominent central process on the hypovalve characterizes this species, but other Peripteropsis species lack processes on their hypovalve. This species may belong to another genus. Chaetoceros sp. in Harwood (1986, pl. 3, fig. 1) is similar to this species because it possesses a central prominent process on the valve, but it may not belong to Peripteropsis because it connected by slender seta which Peripteropsis species lack. This species may belong to another genus.

DISCUSSION
The morphology, taxonomy and biostratigraphy of the fossil resting spore morphogenus Peripteropsis formerly known as 'Periptera' are described in this paper. The genus Peripteropsis is characterized by elongated processes at the valve shoulder. Of the four species in this genus, Peripteropsis tetracladia and P. norwegica are similar to the fossil resting spore genera Monocladia, Dicladia and Syndendrium in possessing dichotomous branching processes on their valves (Suto 2003a), suggesting a close phylogenetic relationship between the four genera. However, Peripteropsis is clearly distinguished by the possession of processes on the valve margin, which are lacking in the other three genera. In addition, Peripteropsis differs from Monocladia, Dicladia and Syndendrium by having a relatively flat valve face in contrast to the strongly domed valve faces of the other three genera. Among the morphogenera of Chaetoceros, no others have such processes on the valve margin, and therefore Peripteropsis is distinct and is clearly characterized by the presence of marginal processes.
The genus Peripteropsis appeared in the early Oligocene and became extinct in the early late Miocene (Fig. 3) in the Norwegian Sea at Site 338. The oldest species P. tetracladia arose in the early Oligocene in the stratigraphic records of the Norwegian Sea and disappeared in the early middle Miocene (Fig. 4). In the late Oligocene, P. norwegica appeared and increased upward in abundance through the early Miocene to middle Miocene (Fig. 3). The last species P. tetracornusa appeared in the earliest middle Miocene and became extinct in the earliest late Miocene. 'Peroptera' petiolata occurs through the early Miocene to the earliest middle Miocene. 'Periptera' schraderi? (see Figs 41,42,49) rarely occurs in the early middle Miocene, but its stratigraphic range is not well known because of its rare abundance.   fig. 4) from the Fur Formation, Denmark, in the late Paleocene and early Eocene. These species possess elongated processes at the valve shoulder, but they do not belong to any Peripteropsis species described in this study because these are characterized by their central processes, not on the valve shoulder. These taxa occurred from the Paleocene and Eocene deposits; therefore they may be the oldest Peripteropsis species. Moreover, Harwood & Gersonde (1990) reported many resting spores, except for Peripteropsis, from the lower Cretaceous sediments in the eastern Weddell Sea, east Antarctica.
Each Peripteropsis species has its own geographic distribution pattern. Peripteropsis tetracornusa occurs abundantly in the northwest and northeast Pacific (Suto 2003b) and is also encountered in the Norwegian Sea, although its abundance there is scarce (Table 1). Thus, P. tetracornusa is probably a cosmopolitan species mainly distributed in the Pacific. Peripteropsis tetracladia and P. norwegica have been noted only in the North Atlantic, and therefore they may be endemic to this ocean, although available geographic distributional data are very limited at present.
Some species of Peripteropsis are biostratigraphically useful in the Norwegian Sea and the North Pacific. Suto (2003b) indicated that the last occurrence of Peripteropsis tetracornusa (c. 10.3 million years before present) may be a distinct biohorizon and useful for identifying the upper part of the Zone NPD 5C in the North Pacific. Peripteropsis tetracladia occurs continuously in the interval from the bottom of the late early Oligocene Pseudodimerogramma filiformis Zone through the early middle Miocene Zone NPD 4A, and therefore the species is useful for identifying this interval.
Many detailed descriptions of extant Chaetoceros vegetative cells have been published in previous works (e.g. Cupp 1943; Rines & Hargraves 1988;Hasle & Syvertsen 1996). On the other hand, our knowledge on morphologies of their resting spores is limited, mainly because it is difficult to observe in detail in valve view in the case of extant resting spores in their frustule of vegetative cells. Therefore, more detailed studies about extant and fossil resting spores morphology are needed. When the morphologies and combinations between resting spores and vegetative cells are clarified, we can identify the same species of fossil resting spores and extant vegetative cells.
The significant role of Chaetoceros in marine primary production and ecology has been known for a very long time (e.g. McQuoid & Hobson 1996;Itakura 2000). The abundance changes of fossil resting spores must be effected by changes in nutrient conditions and/or water-column stratification. Therefore the fossil resting spores abundance changes may indicate the past fluctuations and be used as a paleoecological tool. However, the past ecological influences to resting spores have been neglected because the taxonomies of fossil resting spores are not completed and their ecological information is scant.
It is significant to classify correctly other fossil resting spores in all upwelling regions and several ages and to investigate how old the genus Chaetoceros is, in order to understand past productivity and to know the age when circumstances changed to suit initiation and ensuing evolution of this genus. Therefore the establishment of a reliable taxonomy for all fossil resting spores will allow us to detect additional paleoceanographic signals recorded in the fossil resting spore assemblages in upwelling regions.