The Development of Nothofagus Seed
(Including a Preliminary Account of the Embryogeny, etc.)

N. obliqua and N. procera. Flowering season of 1948–49; pistillate flowers were collected about the time of anthesis (middle of October) from trees growing in the garden of B. C. Aston, Wellington, and the grounds of the Plant Diseases Division, D.S.I.R., Auckland. In December, ovules were obtained from developing nuts collected by G. H. Hocking from Kiwitea, Feilding.

Fagus sylvatica. Flowering season of 1948–49; pistillate flowers were collected at irregular intervals from the time of anthesis at the beginning of October until late December from a tree growing in Wellington.

Material was fixed in formalin acetic alcohol. Preservation of complete flowers was made until two or three weeks after anthesis. By this time lignified tissue had begun to form in the developing nut so that it was necessary to take out the developing ovules and to preserve them separately. Embedding was done by the tertiary butyl alcohol method described by Johansen (1940). Sections were cut with a Cambridge rocker microtome usually to thicknesses of 12–16μ. Stains used were safranin and fast green, Delafield's haematoxylin (both Johansen, loc. cit.) and safranin and analine blue. Haematoxylin proved to be a most suitable stain generally and was the one finally used for staining most of the sections.

Development Leading to Seed Formation

The trimerous nuts of Nothofagus develop three and the dimerous nuts two loculi, each loculus having two pendulous ovules which finally become anatropous. The loculi of an individual nut are not clearly separated and might strictly be considered a single loculus. The young ovary consists of an undifferentiated mass of parenchyma. The first appearance of change is the enlargement about the centre of the ovary, and in both transverse and longitudinal sections, of two adjacent rows of cells—in a bimerous flower (fig. 3). A space develops between these two rows. Then, in transverse section, young ovules appear at each end of the rows of cells and on either side of the space (figs. 4 and 5). The ovules then grow by quick division of the cells, forming bulges. In longitudinal section each end of the space, or in other words the developing locules, appears half-moon shaped with the bulge to the outside of the ovary and the developing ovules on the axile side (figs. 1 and 2).

About the time of pollination the slits only, or the first stages in the formation of the ovules are showing.

Ovules develop slowly after pollination. They become pendulous from the roof of the loculus or the top of the placenta, which consists of tissue remaining between the developed loculi. Before fertilization, which occurs some nine to ten weeks after pollination, the ovules become anatropous and develop a single integument (figs. 8 and 9). After fertilization all except one ovule in a nut abort; it is not known whether the ovule which does develop is the first one to be fertilized or not. The divisions between the three loculi finally break down as the single seed develops. This seed grows to occupy the whole internal space of the nut, thus forming an exalbuminous one-seeded nut, with the vascular strands of the placenta lying along one side. The aborted ovules shrivel and almost disappear.

Duration of Stages in Seed Development

The following table gives a comparison of the times taken, where known, for the seed of the species studied to reach certain stages of development,

Details of Developmental Anatomy and Embryogeny

The description will be based on the stages examined in N. menziesii, since the most complete series was obtained for that species.

Anthesis of the flowers on the one tree occurred over a short period. A few weeks before anthesis the ovary consisted of a mass of undifferentiated parenchyma, but at anthesis itself the beginning of loculus and ovule formation was to be seen. The first stages of ovule formation were as described in the section relating to the general sequence of seed development.

It is noted that Langdon (1939), describing the early stages in the development of Fagus americana pistillate flowers writes, “… the portion of the floral axis inclosed by the carpels produces a short, thick columnar structure, at the tip of which ovule development is initiated. Centripetal growth of the incurved carpel margins takes place concurrently with development of the placental axis, and as cushion-like masses of parenchyma they envelop the placentae and ovules.” No such sequence occurred in the species of Nothofagus examined, the loculi and ovules developing from an undifferentiated mass of parenchyma. It is possible that the undifferentiated mass might have been formed at an earlier stage by invagination; a sufficient series of material was not examined to determine this. But not long before anthesis this tissue was completely uniform.

Benson (1894) gives no account of the formation of ovules in Fagus sylvatica, but she did find that they were formed about a month before pollination and that the rudiments of the inner integuments were to be seen at that time. The nucellus and archesporial cells had also begun to form. Langdon found that in Fagus americana, “the formation of ovules takes place before the opening of the buds and the emergence of the pistillate heads.”

In the flowers of the Nothofagus species examined the development of the ovules was all of the same pattern, and at anthesis ovules were usually just beginning to appear, though in occasional flowers the ovary was still undifferentiated. Flowers examined from local Fagus sylvatica had ovules well developed at anthesis.

About four to five weeks after anthesis ovules were well developed and anatropous, and showed a single integument surrounding the nucellus (figs. 8 and 9). Single integuments only were seen in all Nothofagus material. According to Benson (loc. cit., t. LXVIII) there are two integuments in F. sylvatica, but I am unable to confirm this.

The fully developed nucellus is six to seven cells wide with an outer layer of regular cells and the inner cells in distinct parallel rows. The central core, about two cells wide, consists of more elongated cells, while at the base of this core, at the junction of the nucellus and integument, the cells are greatly elongated and more densely staining (fig. 11). The integument forms a narrow micropyle opening towards the base of the style.

The megaspore mother cell arises from one of the top cells of the central row of elongated cells (fig. 10). In one nucellus there was a suggestion of a linear tetrad of four megaspores arising from the megaspore mother cell (fig. 11). These were large cells, while the nucellar cells around them had degenerated.

Embryo Sac

From the slides examined it seems that the chalazal megaspore* enlarges to form the functional megaspore, while the remaining megaspores and surrounding nucellar cells degenerate and form a densely staining tissue around the sac. At the two-nucleate stage the sac becomes vacuolate, with one nucleus at the micropyle and one at the chalazal end (fig. 12). At the four-nucleate stage a pair of nuclei surrounded by cytoplasm is situated at either end of the sac with a marked vacuole in between (fig. 13). With further division of the nuclei the sac becomes much enlarged, the nucellar tissue disappears, and the sac comes in direct contact with the inner layer of the integument.

The final organization of the sac is normal, with an egg apparatus of two densely staining synergids, and the egg, all situated at the micropylar end, three antipodal cells, and a large secondary nucleus in the centre of the sac (fig. 14). Later this nucleus takes up a position on the sac wall. In N. fusca and N. truncata the egg apparatus, before fertilization took place, moved down the wall of the sac a short distance from the micropylar end (fig. 18).

A distinctive feature is that at the antipodal end of the sac the basal elongated nucellar cells remain jutting into the sac as a “column” of tissue (fig. 17). The antipodal cells are situated on top of this (fig. 19). In N. menziesii ovules this column gradually disappears and at the time of, or just after, fertilization is not to be seen, but in N. fusca ovules it remains for some time as a cone-shaped body of cells (fig. 18). This type of structure was first described and named “Postament”—an old German word for base, pedestal, or stand—by von Westermaier (1890). Dahlgren (1940) gives an account of it in a number of families. Of the two possible methods of origin, a resistant remnant of the nucellar tissue or a special growth of tissue from the chalazal end of the embryo sac, Dahlgren favours the former. As far as could be ascertained in Nothofagus, the Postament consists of a resistant part of the nucellus. Dahlgren mentions the elongation sometimes of the cells of the Postament and advances this and other evidence in support of a contention that it serves the function of conducting tissue. In Nothofagus this is possibly so, because the basal cells of the Postament were connected with the vascular tissue leading from placenta to ovule and cytoplasm surrounded the Postament.

A flat-topped Postament is also seen present in the ovules of N. obliqua and N. procera (fig. 17). In both these species some of the basal cells of the column have very thick, dark-staining walls.

* The terminology of Maheshwari (1948) is used.

Describing the embryo-sac formation in F. sylvatica ovules, Benson (1894) stated that central strands of the nucellus arising from a layer of sub-epidermal cells maintained regular periclinal division. They stood out from the surrounding tissue and formed the “sporogenous tissue.” Some of the cells of the central core elongated without dividing, and it was suggested that one of these long cells formed the embryo sac. The embryo sac had in any case an inconspicuous beginning and “there is no other mark by which to distinguish it from its sister cells above and below it in an axile row than the formation of the two nuclei which take up their position at either end of the cell.”

Abortion of Ovules

Some of the ovules in the one nut did not form embryo sacs, though it was usual to find three to four ovules, in a six-ovulate nut, with sacs. The process of fertilization was not seen, but about the time it should have taken place many of the ovules with fully developed sacs began to abort. In the 1947–48 season, a sparse flowering year for beech in the areas from which the material was collected, most ovules aborted at this stage and it was therefore difficult to find any pro-embryos or embryos. In the 1948–49 season it was usually possible to find one ovule per nut with a pro-embryo or embryo, but the remaining ovules aborted. Amongst other members of the Fagaceae, Quercus may have all the ovules fertilized and any one is capable of forming the seed (Busgen, 1916).

In the process of abortion the embryo sac and any nucellar tissue collapsed leaving the integument intact at first. The Postament appeared to be the last tissue to break down, while the secondary nucleus could be seen until quite late stages of abortion. In the final stages of breakdown the chalazal end of the ovule, together with the strands of vascular tissue attaching the ovule to the placental conducting tissue, could be seen lying on the side of the developing seed where this was present.

Embryogeny and Endosperm Development

Although the process of fertilization was not seen, stages immediately following it were obtained. The youngest pro-embryo was at the three-celled stage. From this it appeared that the first division of the zygote had been transverse, while the next division was a longitudinal one of the basal cell (fig. 20). Irregular division of the apical cells then produced an elongated and finally obovate proembryo without any signs of a true suspensor, but with large basal cells (figs. 21 and 22). Further division produced a globose embryo still without an apparent suspensor, but with light-staining basal cells which may have been the original ones (fig. 23).

An unusual feature, for this family, was the slow commencement of the endosperm; the proembryo had reached a multiple-celled stage before there was any appearance of it. When it did form, the free-nucleate stage was followed quickly by large-celled endosperm with sparse strands of cytoplasm and large nuclei. Endosperm never seemed to fill the entire sac.

Advanced stages of N. menziesii embryos were not seen, but in N. fusca welldeveloped embryos occupied the centre of the sac. They were composed of small cells containing many fat globules. From a central body of tissue two cotyledons were directed towards the micropylar end of the sac and a broad radicle was directed towards the antipodal end (fig. 25).