J. Cell Sci. 72, 307-319 (1984) 307Printed in Great Britain © The Company of Biologists Limited 1984

ULTRASTRUCTURE OF THE FUSION CELL IN

FAUCHEOCOLAX ATTENUATA SETCH.

(RHODOPHYTA, RHODYMENIALES)

STYLIANOS G. DELIVOPOULOS* AND PAUL KUGRENSDepartment of Botany, Colorado State University,Fort Collins, Colorado 80523, U.SA.

SUMMARY

The fusion cell in Faucheocolax attenuata Setch. is a highly lobed, thick-walled, multinucleateand irregularly shaped cell originating from the basal cell of the auxiliary cell branch. The formationof the fusion cell occurs by an incorporation of vegetative cells into the basal cell, after dissolutionof septal plugs between these cell types. Thus the fusion cell is a syncytium containing only haploidnuclei, as well as unusual mitochondria and plastids. Mitochondria lack cristae and instead containa tubular helical structure. Plastids are atypical with regard to thylakoid organization in red algae,because they lack the peripheral thylakoid and their photosynthetic thylakoids are aggregated to oneside. In addition, they contain large osmiophilic bodies. Nuclear envelopes appear to produce largequantities of membrane cisternae. Floridean starch is absent and the cytoplasm contains fewribosomes. The plasma membrane is irregular and endoplasmic reticulum cisternae are situatedparallel to it. Bundles of putative microfilaments were commonly found in nuclei and the cytoplasm.Structural evidence does not support any meristematic, nutritive or secretory functions previouslyascribed to fusion cells in other genera.

INTRODUCTION

Following fertilization, Florideophycean red algae form a diploid phase termed thecarposporophyte. This stage remains attached to the female gametophyte and consistsof three different cell types: the fusion cell, gonimoblast cells and carpospores. Themost unusual component is the fusion cell, which is generally considered to be animportant structure, because apparently it can give rise to gonimoblast filaments(Sjostedt, 1926; Oza, 1976; Kugrens, 1982; Delivopoulos & Tsekos, 1983). Thefusion cell reportedly is formed from repeated fusions of diploid and haploid cells(Wetherbee, 1980; Kugrens & Arif, 1981; Kugrens, 1982; Ramm-Anderson &Wetherbee, 1982; Delivopoulos & Tsekos, 1983), resulting in a large, irregularlyshaped syncytium.

Despite its significance and general occurrence in the Class Florideophyceae, thefusion cell remains one of the least investigated structures of the carposporophyte.Most ultrastructural studies of this cell have been incidental in papers describingcarposporogenesis (Kugrens & West, 1973, 1974; Kugrens, 1982; Ramm-Anderson& Wetherbee, 1982) and with two exceptions (Kugrens, 1982; Ramm-Anderson &

•Present address for correspondence: Botanical Institute, University of Thessaloniki, 54006Thessaloniki, Greece.Key words: Faucheocolax, fusion cell, red algae, ultrastructure.

308 5. G. Delivopoulos and P. Kugrens

Wetherbee, 1982) they have been confined to members of the Ceramiales (Kugrens& West, 1973, 1974; Wetherbee, West & Wynne, 1974; Tripodi & De Masi, 1977,1978; Wetherbee, 1980; Kugrens & Arif, 1981). As a result of this taxonomic selectiv-ity, the extent of variation in structure and function of fusion cells among other ordersof red algae remains enigmatic. To date only three papers have provided detaileddescriptions of the fine structure of fusion cells (Tripodi & De Masi, 1977, 1978;Kugrens & Arif, 1981) and only two considered the fusion cell as a developmentalsystem (Wetherbee, 1980; Kugrens & Arif, 1981).

There remain numerous unresolved problems and ambiguities regarding the ultra-structure of fusion cells. These include: (1) fusion cell function; (2) the process andreason(s) for vegetative and gonimoblast cell incorporation into the fusion cell; (3) thefate of haploid nuclei within fusion cells; (4) the functional significance of differentorganelles and/or structures existing within fusion cells; (5) the presence of anyunusual fusion cell structures; and (6) the timing and extent of fusion cell degenera-tion that often takes place.

Therefore, studies onFaucheocolax fusion cells were initiated to provide additionalinformation in a representative of the Rhodymeniales, Faucheocolax attenuataSetch. The information from this study provides some interesting contrasts andcomparisons with the descriptions of fusion cells from Ceramialean genera (Tripodi& De Masi, 1977, 1978; Wetherbee, 1980; Kugrens & Arif, 1981).

MATERIALS AND METHODS

Thalli of Faucheocolax attenuata bearing cystocarps of varying sizes, were collected near SalmonBank, San Juan Island, Washington and by SCUBA diving north of the Bodega MarineLaboratories, Sonoma County, California. The parasitic thalli were excised (rom Fauchea and fixedimmediately for light and electron microscopy according to previously described procedures(Kugrens, 1974). Thin sections were examined with either an AEI-6B or AEI-801 electronmicroscope.

OBSERVATIONS

Faucheocolax is a parasitic red alga that is related to its host Fauchea (Kylin, 1956)and this is considered an adelphoparasite. The thallus is reduced in size and has a lightpink coloration. The carposporophyte of Faucheocolax appears typical of othergenera in the Order Rhodymeniales of the Florideophyceae, because it is subtendedby a large fusion cell (Kylin, 1956). Fig. 1 summarizes the post-fertilization eventsleading to the formation of this unusual fusion cell, which originates from the basalcell of a two-celled auxiliary cell branch. Subsequently, adjacent cells become incor-porated to form a larger, more irregularly shaped fusion cell. Incorporated cellsalways include vegetative cells and sometimes carpogonial branch cells, as well as thesupporting cell, but never the auxiliary cell. Therefore, only haploid nuclei occur inthis fusion cell.

From the earliest to the oldest developmental stages, the nature of the fusion cellremains unchanged, except that additional incorporation of vegetative cells increasesits size. Even before any cell incorporation is initiated, the fusion cell precursor (the

310 5. G. Delivopoulos and P. Kugrens•J-0*-

AC

Figs 2 and 3

Faucheocolax/usi'ott cell 311

of development. The plug is dark staining and has a mottled appearance. In addition,it exhibits an irregular or crenulate profile on the side of the fusion cell, whereas it issmooth on the side toward the auxiliary cell.

Fusion cell nuclei usually occur in proximity to each other (Figs 2, 3) and aregenerally situated in the peripheral cytoplasm (Fig. 4), often in contact with theplasma membrane or the endoplasmic reticulum cisternae beneath the plasmamembrane (Fig. 4). Furthermore, the nuclear envelopes in newly incorporated cellsgenerally become irregular and highly lobed (Fig. 4). Numerous stacked membranecisternae are characteristically associated with nuclei (Figs 2-4), their appearancesuggesting that they are derived from the nuclear envelope. These membranes areparallel to each other in longitudinal view but appear reticulate in cross-section (Figs3,4).

One of the more unusual structures in these fusion cells are the clusters of smallparallel rods having dimensions and aggregations reminiscent of microfilaments.They are tentatively designated as such. These microfilaments occur in both thecytoplasm (Fig. 5) and nuclei (Figs 4, 6).

Fusion cell plastids are atypical of red algae (Fig. 7). The thylakoids are aggregatedto one side and large osmiophilic bodies are found in the stroma. The peripheralthylakoid is absent.

Mitochondria, likewise, are atypical (Fig. 8). Instead of possessing identifiablecristae, most either lack any internal membranes or they contain a tubular helicalmembrane that is located in the centre of the stroma.

The incorporation of vegetative cells into the fusion cell requires the dissolution ofseptal plugs between the respective cells. Fig. 9 shows the features of the auxiliary-fusion cell septal plug, which appears to be undergoing some degradation on thefusion cell side, as indicated by material being removed from the plug. This plug,however, is never degraded completely, so the auxiliary cell retains its integritythroughout development.

Septal plugs between fusion and vegetative cells are smaller (Fig. 10) and readilyundergo dissolution. Even before the actual fusion is accomplished, the adjacent cellsbegin to deposit wall material, forming a thick wall (Fig. 10). The vegetative cellseptal plugs have densely staining peripheries with a less-dense core (Figs 10, 11).Plug degradation generally is initiated on the fusion cell side and the first indicationof degeneration is the swelling of the septal plug on that side (Figs 11, 12).Microtubules are in the proximity of the swollen plug (Fig. 12). With continuedswelling, the plug eventually breaks down into smaller dark-staining pieces (Figs 13,14), until the entire plug has been fragmented (Fig. 15). The septal plug fragmentsare often associated with a reticulate membrane system, possibly endoplasmic

Fig. 2. Overview of the fusion cell. Numerous fused vegetative cells are incorporated, asindicated by wall remnants (arrows). Note the absence of floridean starch and the paucityof organelles, as well as the disposition of nuclei and organelles toward the peripheralcytoplasm. A thickened wall surrounds the components of the fusion cell. X2950.Fig. 3. The fusion and auxiliary cells connected by a large basal septal plug (sp). Note theirregular outline of the plug on the fusion cell side. X7750.

312 S. G. Delivopoulos and P. Kugrens

Faucheocolax/ust'on cell 313

reticulum (Figs 14-16), that may have been formed by the nuclear envelopes.Cytoplasmic continuity is now established between the cells that contain numerousdark-staining masses of material, probably representing septal plug remnants (Fig.16). Thus the fusion cell now has increased its size and numbers of nuclei, all of whichare haploid. This incorporation of nuclei through cytoplasmic fusions creates asyncytium.

DISCUSSION

Fusion-cell formation and carposporophyte development in Faucheocolax do notmatch the descriptions of Setchell (1923) and Kylin (1956), although the diagrams ofthe mature carposporophyte are similar. Both state that post-fertilization develop-ment in Faucheocolax is identical to Fauchea but details of this development wereprovided only lor Fauchea. We must, therefore, assume that either the descriptionswere erroneous or we are dealing with a different species from the one described bySetchell (1923). This contradiction points out the necessity for a critical re-evaluationof fusion cells in different taxa, so that we can correctly assess the roles assigned tothem by light microscopists.

Studies on Polysiphonia novae-angliae Taylor (Wetherbee, 1980), Asterocolaxgardneri Setch. (Kugrens & Arif, 1981) and Leachiella pacifica (Kugrens, 1982)demonstrated that fusion cells change during carposporophyte development.Therefore, all developmental stages should be examined, so that accurate interpreta-tions of fusion cell structure and function can be made (Brawley & Wetherbee, 1981).This is particularly true in the light of some reports that indicated that fusion cellscontained unusual organelles (Tripodi & De Masi, 1977, 1978) or appeared to bedegenerating (Kugrens & West, 1974). Some important functions are usuallyascribed to the fusion cell (Yamanouchi, 1906; Drew, 1954; Fritsch, 1965; Wether-bee, 1979, 1980; Kugrens & Arif, 1981; Kugrens, 1982), but apparently some of theideas regarding this cell must be modified, on the basis of our findings in Faucheo-colax.

The fusion cells in some species may not be the important components of thecarposporophyte that previous investigators had presumed. Rather, this cell in someof these organisms appears degenerate and seemingly does not play any role in carpos-porophyte development, thus merely representing a vestigial cell. The degradativeenzymes that may form in this cell might initiate fusions with adjacent cells, also

Fig. 4. Fusion cell nucleus showing the apparent production of numerous membranesfrom the nuclear envelope. An expansion of these membranes forms a reticulate networkof endoplasmic reticulum. Note the intranuclear microfilaments (arrows). X16000.

Fig. 5. Cytoplasmic microfilaments in the fusion cell. X 40 000.

Fig. 6. Higher magnification of intranuclear microfilament cluster. X70450.

Fig. 7. Degenerating plastids and proplastids in the fusion cell. X15 000.

Fig. 8. Typical fusion cell mitochondria. Tubular helical membranes (arrows) occur inthe stroma. X 32 150.

5. G. Delivopoulos and P. Kugrens

iFigs 9-16

Faucheocolax fusion cell 315

causing them to assume a similar degenerate appearance. Cytoplasmic fusions occurthrough the dissolution of septal plugs between the fusion cell and adjacent cells,thereby increasing the cytoplasmic volume of the fusion cell and giving it moreprominence in the carposporophytic system than is justified through light-microscopic observations.

Fusion cells in other genera, regardless of their intial functions, apparently alsodegenerate with an accompanying incorporation of gonimoblast and vegetative cells.In these fusion cells, however, degeneration begins only after the carposporophyte iswell formed. On the other hand, Faucheocolax fusion cells begin degeneration duringthe earliest stages of carposporophyte development and never participate in formingany products of the carposporophyte. Faucheocolax gonimoblast cells apparently areprotected from incorporation by the large, basal septal plug between the auxiliary celland the fusion cell. This plug shows some degeneration on the fusion cell side but acytoplasmic connection is never created. In addition, the auxiliary cell could act as abuffer cell, intervening between the meristematic gonimoblast cells and the fusioncell.

Since the multinucleate condition of fusion cells arises from the fusions of haploidand diploid cells (Wetherbee, 1980; Kugrens & Arif, 1981; Kugrens, 1982; Ramm-Anderson& Wetherbee, 1982; Delivopoulos&Tsekos, 1983), it is assumed that thereis some means of isolating or inactivating the haploid nuclei (Kugrens & Arif, 1981). Todate the only isolating mechanism reported is a cytoplasmic barrier (Kugrens & Arif,1981), which thus far is unique for the Rhodomelaceae. Obviously there is no need fornuclear degeneration or inactivation in Faucheocolax fusion cells, since only haploidnuclei are present. Nevertheless, nuclear envelopes seem to produce large numbers ofmembranes, which usually aggregate into stacks. These membranes often completelysurround the nuclei and are sometimes associated with septal plag fragments. Similar

Fig. 9. Portion of the large septal plug between the fusion and the auxiliary cell. Note thecrenulated profile of the plug on the fusion cell side and the smooth outline on the auxiliarycell side. Dark staining masses (arrows), representing material being removed from theplug, are present in the fusion cell close to the plug, as well as on its surface. X 15 700.

Fig. 10. Young fusion cell connected to two intact vegetative cells. The septal plugsbetween the fusion cell and vegetative cells are in various stages of dissolution. X4570.

Fig. 11. Septal plug showing incipient dissolution indicated by plug enlargement in thefusion cell side. X 13 750.

Fig. 12. Septal plug exhibiting additional dissolution. Microtubules are present near theplug. X18 400.

Fig. 13. Septal plug showing further degeneration. X11250.

Fig. 14. Higher magnification of an advanced stage of septal plug dissolution. Note thereticulate endoplasmic reticulum near the septal plug. X25 000.

Fig. 15. Final stage of septal plug dissolution. The plug remnants occur to one side of thecytoplasmic channel that has been created. The vegetative cell now takes on the charac-teristics of the fusion cell. X 16875.

Fig. 16. Fusion cell and its newly incorporated vegetative cell. Masses of septal plugmaterial are dispersed in the cytoplasm. Arrows indicate wall remnants from fused cells.X11785.

316 S. G. Delivopoulos and P. Kugrens

stacking of endoplasmic reticulum (ER) has been reported in developing and/ormature sieve elements (Melaragno & Walsh, 1976; Singh, 1980). In these instancesit was suggested that the stacked ER may participate in the autolytic process byproviding enzymes for this activity (Esau & Gill, 1972, 1973). The localization of acidphosphatase (Catesson, 1973; Esau & Charvat, 1975) and ATPase (Gilder & Cron-shaw, 1973a,6) on stacked ER support such a view. Although it is difficult to identifythe observed features with certainty, it is possible that a similar function could beascribed to the membranes of Faucheocolax fusion cells, since these cells have someadditional features in common with differentiating and/or mature sieve elements,such as the absence of dictyosomes and ribosomes (Esau, 1977).

The study of Tripodi & De Masi (1977) onErythrocystis tnontagnei (Derb. et Sol.)Silva fusion cells revealed numerous 'unusual or peculiar' structures, similar to theones in Faucheocolax. These included unusual nuclei, mitochondria with helicalinternal membranes, dense cytoplasmic bodies surrounded by endoplasmicreticulum, crystalline endoplasmic reticulum, large concentrations of membranes andsheets of electron-transparent material. One of the most unusual, however, was acylindrical structure consisting of coiled sheets of electron-dense material that inlongitudinal section vaguely resembled a basal body, and in fact was interpreted byTripodi & De Masi (1978) as 'a vestige of a flagellum'. However, it does not actuallyresemble any structure that can be construed as part of a basal body, because thereare no microtubular remnants. A serious effort was made to find these or similarstructures in Faucheocolax, but we failed to locate them.

Plastids of the Faucheocolax fusion cell, as those of Erythrocystis (Tripodi & DeMasi, 1977), lack a peripheral thylakoid and exhibit a degenerate appearance.Furthermore, the twisted or helical membranes in mitochondria, previously inter-preted as mitochondrial DNA (Tripodi, Pizzolongo & Giannattasio, 1972; Tripodi,1974; Tripodi & De Masi, 1977; Tsekos, 1983), are identical to the helical membranesfound in degenerating mitochondria of the Faucheocolax fusion cell. Therefore, webelieve that these structures represent disorganized membranes in a degeneratingmitochondrion.

These previous studies apparently were based on late ontogenetic stages in fusioncell development; it is therefore difficult to ascribe functional significance to many oftheir observed features. Consequently, it seems plausible that the organelles arebecoming disorganized in an aging fusion cell, since similarly structured organellesand membranes were also found in Faucheocolax fusion cells, which we interpretedas aging or degenerating cells.

Plant cells have cytoskeletons composed of actin microfilaments and microtubulesthat are indistinguishable from those found in animal cells (Alberts et al. 1983). Theoccurrence of microfilaments in plant cells and their possible function have beendiscussed by Parthasarathy & Miihlethaler (1972) and reviewed recently by Jackson(1982). In another investigation (Parthasarathy, 1974) microfilament bundles wereobserved in differentiating sieve elements whose protoplasts undergo a profoundchange during ontogeny (Esau, 1977). Electron-microscopic studies of cells under-going changes in shape have shown that microtubules and microfilaments are involved

Faucheocolax fusion cell 317

in these processes (Wessels, 1971). In addition, formation of the actin filaments in thefertilization tubule of Chlamydomonas reinhardii Dan. is clearly dependent uponreceipt of intracellular signals generated by adhesion of flagellar membranes(Goodenough, 1977; Solter & Gibor, 1977). Since the fusion cell in Faucheocolaxchanges its shape rapidly and expands outwardly by fusions with adjacent cells, itseems possible that the cytoplasmic microfilaments are somehow involved in cellularfusions. Similar structures were found in nuclei of Porphyridium purpureum (Bory)Drew et Ross (Bronchart & Demoulin, 1977) and Rhodella reticulata (Deason, Butler& Rhyne, 1983), as well as in Faucheocolax nuclei, but the role of these intranuclearmicrofilaments remains to be demonstrated. Whether cytoplasmic and intranuclearmicrofilaments of Faucheocolax fusion cells are composed of actin needs furtherinvestigation.

A generally held view (Yamanouchi, 1906; Drew, 1954; Fritsch, 1965; Wetherbee,1979, 1980; Kugrens, 1982) is that the fusion cell and any cells incorporated into thefusion cell provide nutrients to the remainder of the developing carposporophyte,since large quantities of floridean starch are found in some fusion cells (Kugrens &Arif, 1981; Kugrens, 1982; Delivopoulous & Tsekos, 1983). Wetherbee (1980)observed that the main body of the Polysiphonia fusion cell was devoid of starch grainsand ribosome concentrations were greatly diminished in that cell. Consequently, itsfunction as a nutrient source for the developing carposporophyte is limited. Neverthe-less, Wetherbee (1980) embraced the idea of a nutritive role for the fusion cell.Meristematic activity (Tripodi & De Masi, 1977; Duckett & Peel, 1978; Kugrens,1982; Delivopoulos & Tsekos, 1983) or secretory functions (Kugrens & Arif, 1981)also were ascribed to the fusion cell. Structural evidence in Faucheocolax, however,does not support a nutritive, secretory or meristematic function for the fusion cell.Thus the fusion cell, while taxonomically important, does not play any active orobvious role in carposporophyte development in Faucheocolax and we suspect thatthe same may apply to other genera where fusion cells are given a prominent role incarposporophyte development.

The lighter and degenerate appearance of the fusion cell is not artifactual or due topoor fixation. Carpospores (Delivopoulos & Kugrens, 1984), as well as auxiliary andgonimoblast cells, were properly preserved and displayed a variety of developmentalfeatures in the carposporophyte. Rather, the Faucheocolax fusion cell has the charac-teristics of a disorganized cell similar in structure to differentiating and/or maturesieve elements (Esau, 1977). Therefore, we conclude that the fusion cell remains alivethroughout carposporophyte development but its function in relation to the femalereproductive system remains enigmatic.

REFERENCES

ALBERTS, C , BRAY, D., LEWIS, J., RAFF, M., ROBERTS, K. & WATSON, J. (1983). Molecular

Biology of the Cell. New York, London: Garland Publishing, Inc.BRAWLEY, S. H. & WETHERBEE, R. (1981). Cytology and Ultrastructure. In The Biology of

Seaweeds (ed. C. S. Lobban & M. J. Wynne), pp. 248-299. Berkeley, Los Angeles: Universityof California Press.

318 S. G. Delivopoulos and P. Kugrens

BRONCHART, R. & DEMOULIN, V. (1977). Unusual mitosis in the red alga Porphyridium pur-pureum. Nature, Land. 268, 80-81.

CATESSON, A.-M. (1973). Observations cytochimiques sur les tubes cable's de quelques angio-spermes.^. Microscopie 16, 95-104.

DEASON, T. R., BUTLER, G. L. & RHYNE, C. (1983). Rhodella reliculata sp. nov., a new coccoidRhodophytan alga (Porphyridiales). J. Phycol. 19, 104-111.

DELIVOPOULOS, S. G. & KUGRENS, P. (1984). Ultrastructure of carposporogenesis in the parasiticred alga Faucheocolax attenuata Setch. (Rhodymeniales, Rhodymeniaceae). Am. J. Bot. 9 (inpress).

DELIVOPOULOS, S. G. & TSEKOS, I. (1983). A light microscopy study of carposporophyte develop-ment in Gracilaria verrucosa (Hudson) Papenfuss. Ann. Bot. 52, 317-323.

DREW, K. M. (1954). The organisation and inter-relationships of the carposporophytes of livingFlorideae. Phytomorphology 4, 55-69.

DUCKETT, J. G. & PEEL, M. C. (1978). The role of transmission electron microscopy in elucidat-ing the taxonomy and phylogeny of the Rhodophyta. In Modern Approaches to the Taxonomyof Red and Brown Algae (ed. D .E .G. Irvine & J. H. Price), pp. 157-204. London: AcademicPress.

ESAU, K. (1977). Anatomy of Seed Plants, second edn. New York, Santa Barbara, London, Syd-ney, Toronto: John Wiley and Sons, Inc.

ESAU, K. & CHARVAT, I. D. (1975). An ultrastructural study of acid phosphatase localization incells of Phaseolus vulgaris phloem by the use of the azo dye method. Tissue & Cell 7, 619—630.

ESAU, K. & GILL, R. H. (1972). Nucleus and endoplasmic reticulum in differentiating rootprotophloem of Nicotiana tabacum.J. Ultrastruct. Res. 41, 160—170.

ESAU, K. & GILL, R. H. (1973). Correlation in differentiation of protophloem sieve elements ofAllium cepa root. J . Ultrastruct. Res. 44, 310-328.

FRITSCH, F. E. (1965). The Structure and Reproduction of the Algae, vol. 2. Cambridge UniversityPress.

GILDER, J. & CRONSHAW, J. (1973a). Adenosine triphosphatase in the phloem of Cucurbita.Planta 110, 189-204.

GILDER, J. & CRONSHAW, J. (19736). The distribution of adenosine triphosphatase activity indifferentiating and mature phloem cells of Nicotiana tabacum and its relationship to phloemtransport. J. Ultrastruct. Res. 44, 338-404.

GOODENOUGH, U. W. (1977). Mating interactions in Chlamydomonas. In Microbial Interactions,Receptors and Recognition, Series (ed. B. J. Reissing), pp. 322-350. London: Chapman & Hall.

JACKSON, W. T. (1982). Actomysin. In The Cytoskeleton in Plant Growth and Development (ed.C. W. Lloyd), pp. 3-29. New York, London: Academic Press.

KUGRENS, P. (1974). Light and electron microscopic studies on the development and liberation ofjfanczewskiagardneri Setch. spermatia (Rhodophyta). Phycologia 13, 295-306.

KUGRENS, P. (1982).Leachiellapacifica, gen. etsp. nov., a new parasitic red alga from Washingtonand California. Am. J. Bot. 69, 306-319.

KUGRENS, P. & ARIF, I. (1981). Ultrastructure of the fusion cell in Asterocolax gardneri. J. Phycol.17, 215-223.

KUGRENS, P. & WEST, J. A. (1973). The ultrastructure of carpospore differentiation in the parasiticred alga Levringiella gardneri (Setch.) Kylin. Phycologia 12, 163-173.

KUGRENS, P. & WEST, J. A. (1974). The ultrastructure of carposporogenesis in the marinehemiparasitic red alga Erythrocystis saccata.J. Phycol. 10, 139-147.

KYLIN, H. (1956). Die Gattungen derRhodophyceen. Lund: CWK Gleerups Forlag.MELARAGNO, J. E. & WALSH, M. A. (1976). Ultrastructural features of developing sieve elements

in Lemma minor L. The protoplast. Am. J. Bot. 63, 1145-1157.OZA, R. M. (1976). Studies on Indian Gracilaria. II. The development of Gracilaria corticata.

Botanicamar. 19, 107-114.PARTHASARATHY, M. V. (1974). Ultrastructure of phloem in palms. I. Immature sieve elements

and parenchymatic elements. Protoplasma 79, 59-91.PARTHASARATHY, M. V. & MOHLETHALER (1972). Cytoplasmic microfilaments in plant cells.

jf. Ultrastruct. Res. 38, 46-72.RAMM-ANDERSON, S. M. & WETHERBEE, R. (1982). Structure and development of the carpos-

porophyte of Nemalion helminthoides (Nemalionales, Rhodophyta).^. Phycol. 18, 133-141.

Faucheocolax fusion cell 319

SETCHELL, W. A. (1923). Parasitic florideae, II. Univ. Calif. Publs Bot. 10, 393-396.SINGH, A. P. (1980). On the ultrastructure and differentiation of the phloem in sugarcane leaves.

Cytologia4S, 1-31.SJOSTEDT, L. G. (1926). Floridean studies. Lands Univ. Asskr. N.F. II22, 1-94.SOLTER, K. M. & GIBOR, A. (1977). Evidence for role of flagella as sensory transducers in mating

of Chlamydomonas reinhardii. Nature, Land. 265, 444—445.TRIPODI, G. (1974). Ultrastructural changes during carpospore formation in the red alga

Polysiphonia J. submicrosc. Cytol. 6, 275-286.TRIPODI, G. & D E MASI, F. (1977). The post-fertilization stages of red algae: the fine structure of

the fusion cell of Erythrocystis J'. submicrosc. Cytol. 9, 389—402.TRIPODI, G. & D E MASI, F. (1978). A possible vestige of a flagellum in the fusion cell of the red

alga Erythrocystis montagnei. jf. submicrosc. Cytol. 10, 435-439.TRIPODI, G., PIZZOLONGO, P. & GIANNATTASIO, M. (1972). A DNase-senaitive twisted structure

in the mitochondrial matrix of Polysiphonia (Rhodophyta). J. Cell Biol. 55, 530—532.TSEKOS, I. (1983). The ultrastructure of carposporogenesis in Gigartina teedii (Roth) Lamour.

(Gigartinales, Rhodophyceae): Gonimoblast cells and carpospores. Flora 174, 191-211.WESSELS, N. K. (1971). How living cells change shape. Scient. Am. 225(4), 77-82.WETHERBEE, R. (1979). 'Transfer connections': Specialized pathways for nutrient translocation in

a red alga? Science 204, 858-859.WETHERBEE, R. (1980). Postfertilization development in the red alga Polysiphonia. 1. Proliferation

of the carposporophyte. J. Ultrastruct. Res. 70, 259-274.WETHERBEE, R., WEST, J. A. & WYNNE, M. J. (1974). Annulate lamellae in post-fertilization

development in Polysiphonia novae-angliae (Rhodophyta). J. Ultrastruct. Res. 49, 401-404.YAMANOUCHI, S. (1906). The life history of Polysiphonia violacea. Bot. Gaz. 41, 425, 433.

(Received 3 April 1984 -Accepted 25 June 1984)