|Constancea 83, 2002|
University and Jepson Herbaria
P.C. Silva Festschrift
This paper is dedicated to Paul C. Silva on the occasion of his 80th birthday. René Delépine is particularly delighted to show his gratitude to Dr. Silva who, together with late Professor J. Feldmann, very kindly gave him advice during his first phycological research work on Codium ...some time ago (!). Bruno de Reviers is very grateful for all the advice kindly provided by Dr. Silva, and is also delighted to dedicate this article to him.
Skottsberg (1907) made the combination Adenocystis utricularis (Bory de Saint-Vincent) Skottsberg, based on Asperococcus utricularis Bory de Saint-Vincent (1825 ) for material collected from the Falkland Islands (as îles Malouines) and Chilean coasts. Later, Skottsberg (1921) provided a more detailed account of this species and described the new form Adenocystis utricularis f. longissima which he considered might be conspecific with Chorda rimosa Montagne, as he wrote: An = Chorda rimosa Mont (The Latin conjunction An could be translated here either isn't? or wouldn't it be?). Chorda rimosa was placed in Scytosiphon by Endlicher (1843: 25) and in Tubicutis, a synonym of Scytosiphon, by Kuntze (1898: 434), these authors referring to Agardh's (1820) concept of Scytosiphon.
Adenocystis utricularis f. utricularis is known from the Falkland Islands, Patagonia, Tierra del Fuego and the Antarctic Peninsula, as well as from the South Orkney Islands, South Shetland Islands, South Georgia, Crozet, Kerguelen, Marion Island, Campbell Island, Auckland Islands, Chatham Islands, Macquarie Island, Stewart Island, southern New Zealand (Lindauer et al. 1961; Papenfuss 1964; Ricker 1987), and southern Tasmania (Cribb 1954; Womersley 1987).
According to Skottsberg (1921), Adenocystis utricularis f. longissima occurs in several localities in the southern South American regions of Patagonia, Tierra del Fuego, and the Falkland Islands. Skottsberg (1941: 37) also mentions this form from South Georgia (Cumberland Bay, 23/iv/1902).
Recently, an unidentified Adenocystis (Adenocystis sp.), was recorded from New Zealand by Nelson (1994) collected from the Bounty Islands as well as from the Antipodes, Auckland, and Campbell islands.
Therefore three taxa are currently included in the genus Adenocystis (1)Adenocystis utricularis f. utricularis with a very large geographical distribution; (2) Adenocystis utricularis f. longissima; (3) Nelson's Adenocystis sp.; the latter two with a much more restricted geographical distribution. Furthermore Chorda rimosa may belong to that genus as suggested by Skottsberg (1921) when he described Adenocystis utricularis f. longissima.
The aim of this study is to assess whether or not Chorda rimosa Montagne belongs to the genus Adenocystis and if so, to ascertain its relationships with the other species of the genus. For this purpose, syntype material of Chorda rimosa Montagne, housed at PC, was studied.
The material of C. rimosa (Figs. 1-3) appears fragmentary with no complete individual since no undamaged apex could be observed. For this reason and since no young thalli were available, apical growth could not be investigated. Fragments appear very elongated in gross morphology. The thallus is regularly cylindrical, only a few millimeters wide in this dry state. We did not rehydrate it in order not to damage further the only material currently known of this taxon. The contorted aspect, emphasized by Montagne (1842, 1845), is illustrated in his original drawing, preserved at PC under no. MS-435, p. 5055 (Figs. 4 and 5), and by Kützing (1858, vol VIII, Tab. 15 - text no. 2108 p. 7). The type material also shows this contortion clearly under the dissecting microscope, along with the tears that look like button-holes in Montagne's drawing Fig. 4 lower right); but the material is not well preserved and is difficult to study. However, transverse (Fig. 6) and longitudinal sections (Figs. 7-8) were made in specimens of the lectotype collection (Arrows, Fig. 3) and observed with a research microscope and TEM.
The thallus is hollow. In transverse section (Fig. 6) several zones were observed: (1) a fertile external region with paraphyses and sporangia; (2) a densely coloured zone formed of 1-3 layers of small isodiametric cells of 5-7 µm in studied specimens; (3) a colorless region formed by 5-8 layers of larger cells; and (4) a very thin region formed by compacted cell-structures, more or less reduced to their walls, bordering the central cavity.
In longitudinal section of adult thalli (Figs. 7-8), zones 1, 2, 3 described in transverse section can be easily recognized, but we can now interpret zone 4 as part of the network of inner longitudinal filaments formed by axially elongated cells which appear as an irregular internal zone in transverse section. These medullar longitudinally elongated filaments are shown by Montagne on an unpublished original drawing (Fig. 5), but the section shown seems to have been made in a young part of the thallus without the mature developed network. The medullar filaments appears in Kützing's plate of C. rimosa (1858, Tab. 15, figs. i, k). It must be pointed out that this network is quite different from the regularly arranged hyphae observed in Chorda (illustrated e.g., by Fritsch (1945: 222, fig I).
Sori of unilocular sporangia interspersed with paraphyses were very frequent in many sections (Figs. 6 and 7-8). Original drawings by Montagne (Fig. 4) of these paraphyses could give the impression that unicellular (white arrowhead) and multicellular (black arrowhead) paraphyses are present. Actually, we observed only unicellular organs, no multicellular ones. However their content appears sometimes fragmented and this is likely what Montagne (Fig 4) illustrated and the reason why he, erroneously, considered them multicellular. This could be also due to fertile epiphytes which are frequent on Adenocystis (Reviers and Delépine 1981). Within cells, TEM sections (Fig. 9) clearly show plastids containing thylakoids (white arrow) and a granular zone without thylakoids interpreted as a pyrenoid (asterisk).
We examined herbarium specimens housed in WELT of this Adenocystis sp., collected at Penguin Island (Bounty Islands) and sent by W. Nelson, (Fig. 10). Gross morphology of these specimens resembles that of Chorda rimosa although they are smaller and do not appear contorted. In transverse sections, we observed the same structure as in C. rimosa with the same succession of different regions (Figs. 11, 12, 13, 14, 15. The vegetative regions of the mature thalli show 1-2 external epidermal layers of small pigmented cells and several cortical layers of larger and colorless cells. Their central region is hollow, surrounded by some vertical filaments formed of long, anastomosed cells, similar to what exists in Adenocystis utricularis and Chorda rimosa. In fertile regions, sporangia interspersed with paraphyses are present. The paraphyses and sporangia characters seem also similar in the two plants.
It is important to emphasize that even small plantlets of 2-3 mm are generally fertile. In these young individuals a transverse section shows an external layer of unicellular and claviform paraphyses with unilocular spororangia, similar to what is observed in Adenocystis utricularis f. utricularis and Chorda rimosa. At this step of development, the central region is filled by longitudinal filaments (Fig. 11) formed of long and anastomosed cells identical to what Delépine and Asensi (1978) described in Adenocystis plantlets, with, sometimes, the beginning of a central cavity (Fig. 12).
Diagnostic characters of species and forms mentioned in this paper are summarized in Table 1. As regularly arranged medullary hyphae are not present in Chorda rimosa this species cannot be included in the genus Chorda. Moreover, paraphyses do not have mucilaginous thickened tops in Chorda (e.g., Kawai et al. 2000) whereas our specimens have paraphyses with mucilaginous thickened tops (Figs. 13, 14). On the other hand, anatomical, cytological, and reproductive structures of Chorda rimosa match Adenocystis quite well in organization and we believe that the Montagne species should be transferred to the latter genus. The formal combination will be proposed elsewhere.
Furthermore, morphological and structural characteristics of Adenocystis rimosa are very similar to those of Nelson's (1994) Adenocystis sp. from the Bounty Islands, except the contorted aspect of Chorda rimosa. However, as pointed out by Fritsch (1945: 223): When a piece of the thallus is allowed to dry, it becomes markedly contorted [. . .].We conclude that the difference in amount of contortion is probably not sufficient to separate the two and postulate that Chorda rimosa Montagne and Adenocystis sp. from New Zealand are conspecific.
|Adenocystis utricularis f. utricularis||Adenocystis utricularis f. longissima||Adenocystis sp. from New Zealand||Chorda rimosa||Genus Chorda|
|Length||10-20 cm||reaching 1 m||5-10 cm||5-10 cm||reaching 8 m|
|Gross morphology||broad utricle||long swollen utricle||narrow regular tube||narrow regular tube||very long, terete whip-like thallus, hollow only in old parts|
|Medulla||network||network||network||network||regularly arranged hyphae|
|Paraphyses with mucilaginous thickened top||yes||yes||yes||yes||no|