Botanical Journal of the Linnean Society, 72: 115‑148. With 8
plates and 3 figures
Reproduced here with the
permission of the author
The floral anatomy of Victoria Schomb.
E. L. SCHNEIDER
OBSERVATIONS - Continued
fruit of Victoria
The change in floral architecture of Victoria at the time of anthesis (cf. Plates 1C and 2A, B) is well known and has been repeatedly discussed in the literature (see Faegri & van der Pijl (1971), Valla & Cirino (1972) and especially Prance & Arias (1975)). Correlated with the structural change is an increase in floral temperature and the concomitant production and diffusion of a fragrant odour which has long been believed to be the attractant for the beetle pollinators. Although many of these various phenomena have been observed in the plants grown at Lotusland, Santa Barbara (Table 1), beetles, which were present in the water, were not observed to be either attracted to or temporarily entrapped in the flowers. These observations, although relatively insignificant, lend slight support to the hypothesis of Valla & Cirino (1972) and the experimental conclusions of Prance & Arias (1975), that under certain conditions self‑pollination can be accomplished.
Although pollen grains were not found in sectioned material, they were observed on the stigmatic surface of both mature flowers and fruits. Under moderate magnification, pollen tubes could be seen entering the region between the appressed ventral carpellary surfaces. This conforms to the pathway observed in Nuphar and Nymphaea (Moseley, 1965, 1961) and closely resembles that found in such Ranalian genera as Drimys (Bailey & Nast, 1943), Tasinannia (Smith, 1969) and Degeneria (Bailey & Smith, 1942).
Following pollination and
subsequent fertilization, several striking structural and developmental changes
occur in the maturation of the fruit. Although, as indicated in Table
1, only twelve fruits were observed, all of which were sectioned by hand or
by the use of a commercial meat slicer, the developmental changes were found in
all specimens examined, and thus can be taken as an indication of general fruit
development for Victoria.
One of the more noticeable changes is one of differential horizontal expansion or growth of the proximal and peripheral dorsal carpellary regions. This is best correlated with the concomitant increase in seed size. In contrast, little, if any, horizontal or vertical expansion occurs in either the central receptacular or the ventral carpellary regions (Plate 2C). The outer, horizontal expansion which does occur is due principally to the increase in existing cell size and to the consequent expansion and formation of aerenchyma tissue. It can be assumed that there is some cellular division, although direct cell counts were not taken to confirm this opinion. There is, however, no increase in the number of major vascular bundles. These veins are, of course, increased in length as expansion occurs; and, since the xylem is composed of extensible tracheary types, discontinuities do not appear.
Concurrent with the horizontal expansion of the dorsal carpellary regions, the floral appendages (i.e. sepals, petals, staminodia and stamens) fall or decay away, leaving only scars to indicate their former positions. In addition, the ventral carpellary surfaces, which were rather loosely appressed at the time of pollination, become tightly appressed so that separation of these surfaces, which was possible prior to fertilization by use of a blunt probe, then becomes difficult without considerable force.
Although the outer ovary wall remains firm during fruit development, it is irregularly forced apart due to both the enlargement of the seeds and the swelling of the mucilate contained in each locule. Dehiscence is therefore, in the broadest sense, loculicidal. In Victoria, the seeds are distributed both individually and passively. This is in contrast to Nuphar (Moseley, 1965), in which the seeds of each locule are released together, enclosed in a packet of aerenchymatous tissue. Conard (1905) reported a similar tendency in Nymphaea flava. In Victoria, as each seed is released from the fruit, it is completely surrounded by the translucent, aerenchymous aril. The aril, by entrapment of gases, aids in the dissemination of each seed. Following the irregular separation and disintegration of the ovary wall, total consummation of the fruit is accomplished by the decay of both the remaining receptacular and carpellary tissues.
The origin of the flower ‑ the
Although the origin of the flowering plants is intriguing, it remains as one of the most troublesome, complex questions in biology. During the past 60 years, however, relatively significant advances have been made in elucidating the riddle. One of the foremost advancements was established by the investigations of Bailey & Tupper (1918). By surveying an extraordinarily large number of both extant and fossil vascular plants, their research established that the major underlying phyletic trend in the secondary xylem elements was toward a reduction in length, and that, among the dicotyledons, the longest xylary elements occurred among the Ranales (sensu lato; sensu Eames, 1961). Subsequent studies (Bailey, 1920; Henes, 1959) established that the same phyletic trend exists in the primary xylem. These investigations are extremely important for they reveal extrinsic unidirectional, anatomical evolutionary trends. In other words, these investigators utilized trends that were initiated in more primitive plants than angiosperms. The use of only intrinsic trends in the angiosperms led to the circular reasoning prevalent in the latter half of the nineteenth century which resulted in Engler & Prantl's system. Following the work of Bailey & Tupper, numerous additional xylary trends were established. The more important ones are (1) that scalariformly pitted tracheary elements were eliminated very early in the phylogeny of the Coniferales, Ephedrales and Gnetales (see Bailey, 1953); and (2) that perforate xylary elements in the dicotyledons arose from scalariformly pitted tracheids (Frost, 1930a) (for monocotyledons, see Cheadle, 1943) such as the type which occur in genera of the Ranales which lack vessels (Carlquist, 1961); and (3) that the scalariform perforation plate is primitive and the simple (uniperforate) plate advanced (Frost, 1930b).
Additional extrinsic trends, concerning features other than the xylem, have also been established: namely, (1) the tendency for sieve‑tube elements to change from those most like gymnospermous sieve cells toward shorter, broader types with a concomitant decrease in sieve plate inclination and increase in the size of the sieve pores (Zahur, 1959; Esau, Cheadle & Gifford, 1953; Esau, 1969); (2) the tendency from 1‑sulcate (colpate) pollen in the gymnosperms and Ranales towards multisulcate and/or nonsulcate types (Erdtman, 1954; see also Walker, 1974) in the dicotyledons. These extrinsic trends are unique in that they have left little doubt that the Ranales sensu lato possess the greatest number of primitive characters of any dicotyledons.
Since the Ranales are considered to be primitive angiosperms by many (e.g Corner, 1963; Cronquist, 1968; Guedes, 1973; Sporne, 1969; Takhtajan, 1959, 1969), it has been repeatedly expressed (Bailey & Swamy, 1951; Moseley, 1965) that investigations of various Ranalian subtaxa may reveal additional primitive angiosperm features which, if discovered and evaluated, may contribute to the elucidation of the angiosperm ancestor(s). Repeated emphasis has, however, been focused on the principle that, if a natural (phyletic) system is to be constructed, it must be founded on a synthesis of evidence taken from all available sources of biological investigations. It was, therefore, the specific aim of this investigation to gather data from the various subdisciplines of floral anatomy (especially vascularization and ontogeny) and utilize these data together with previously accumulated data to increase our understanding of (1) the origin and nature of the angiosperm flower, (2) the relationships among the nine genera of the Nymphaeaceae sensu lato, and of the Nymphaeaceae to other families, and (3) to gain a better understanding of floral anatomy in general and of the Nymphaeaceae specifically.
The Ranales‑floral anatomy
The various topics relative to floral anatomy (e.g. gross morphology, pollination and dispersal mechanisms, ontogeny, teratology, histology and vascularizati6n, and the role of each in the solution of morphological problems) have been the subject of debate since the time of de Candolle's Theorie elementaire de la botanique (1813) and Organographie vegetale (1827). The usefulness of each of these disciplines has been established with varying degrees of certainty, although few have been as convincing as the extrinsic trends established for xylem and phloern (length and structure).
Various authors have taken advantage of the doubts and disagreements surrounding each subject and have presented alternative theories to explain the seeming discontinuities which exist. Recently, Carlquist (1969) published a comprehensive critique concerning floral anatomy. In his critique, which was intended to spark debate, Carlquist deprecates the adduction of such anatomical arguments on the grounds that adaptive modifications to functional requirements have had a greater effect on floral evolution. Although the criticisms put forth by Carlquist are, in several cases, justifiable and/or have been the subject of debate (Eyde, 1971; Puri, 1971; Kaplan, 1971; Moseley, 1972; Schmid, 1972; Gottsberger, 1974), this writer would like to amplify certain ideas expressed by these anatomists, for he feels that certain points made by them are indicators "toward acceptable evolutionary interpretations ... (Carlquist, 1969). The first point, in fact the main point, is one stressed by Moseley (1972) concerning vascularization of the flower, viz. vascular conservatism. Basically, Moseley states that (1) the ultimate use of floral anatomy is one of interpretations and (2) that the (floral) vascular system (like any other discipline!) should be used only as one character among many in taxonomic‑phylogenetic studies. To this should be added, ad nauseam if necessary, that topics such as vascularization or ontogeny are far too valuable sources of evidence not to be considered; for many examples have been clearly shown. where vascularization or ontogeny has revealed otherwise obscure or unrecognizable evolutionary relationships. Another point (which is the main theme of Carlquist's critique) concerns the lack of attention which anatomists have given to the adaptive significance of changes in floral structure. In part, this writer is in accord with this view, but believes that the study of functional anatomy (i.e. adaptations and the biological roles of features) is not a suitable starting point for establishing relationships. The concepts derived from such studies do, however, aid in solving obscure taxonomic‑phylogenetic studies by corroborating such relationships and, of course, making them biologically more meaningful.