Some aspects of reproductive behaviour in Eucalyptus grandis (Hill) Maiden

Abstract

Phenological and morphological studies of the flower were done at Zomerkomst Forest Research Station, in order to obtain a clearer understanding of floral biology. The breeding system and related subjects could then be investigated, with a view to ascertaining the amount of natural selfing, the extent of inbreeding depression and their possible effects on the production of improved seed. The first flowers appear on trees aged two to three years. The flowering season varies widely with altitude; it is mainly from February to June, although some flowering is liable to occur throughout the year. The flowering shoot ends in an active vegetative bud and consists of up to 14 "umbels", each of which is typically seven-flowered and is at first enclosed in six or more bracts. The receptacle of the flower bud bears on its rim two independently shed opercula and a ring of numerous inflexed stamens and it encloses the ovary. By the time the floral buds had become recognisable as such, the sepaline operculum had already developed and the petal primordia had been initiated at their base. Stamina! primordia are formed some ten days later and tetrad formation occurs before the outer operculum is shed. The style is initiated at about the same time as the stamina! primordia. The ovary contains numerous ovules and infertile ovulodes, which are initiated as the time for bract shed approaches. There are two types of ovulodes, which give rise to two types of chaff in the mature fruit. The ovules are anatropous to hemitropous and have two integuments, each two cell layers in thickness. These develop at about the same time that the microspore tetrads are formed. The ovary is inferior, typically with five locules and axile placentation, except at the top where the partitions between the locules are incomplete. The flowers are regarded as being protandrous, or at least partially so, and mainly insect pollinated, especially by honey bees. Pollen shed was mostly complete by the second day after anthesis, and some seed was obtained after cross-pollination done at anthesis, although germination of pollen on the stigma did not usually begin until two or more days later. Maximum seed yield was obtained from pollination done on the fifth day after anthesis, before the stigma developed a distinctly swollen appearance. The stigma remained receptive up to the seventh day after anthesis or later, and there were pronounced day to day changes in receptivity. The fruit is a capsule which is mature in five to seven months, depending on the site. In most clones tested, some fruit was set parthenocarpically. Most seeds may be separated from the chaff on the basis of colour and shape, but not size, although by passing through a sieve of gauge 500 μm, most of the seed was retained. Such seed was the best as regards early vigour and survival and the sieving gave increased purity. In tests of self-fertility, in which all pollinations were done at comparable stages of receptivity, the seed yield from self-pollinated flowers varied in different clones from 2 % to 4 7 % of that from cross-pollination. Out of 45 clones which were self-pollinated, 44 yielded some seed, varying in number up to 24 seeds per capsule. Among the progeny, 15 different types of abnormality were found and four of these were used as the main means of assessing natural selfing. This was commonly about 30 %, but was 83 % in one test of out of season flowering. Mean tree heights gave promise as a possible alternative basis to be used in estimating selfing. Although selfing was rather high in some clones, the species was mainly outbreeding. This conclusion is supported by the consistent inbreeding effect on mean tree heights, which was observed at age 11 to 18 months after sowing the seed. This varied in different clones from 8 % to 49 % comparing self-pollinated progeny with that from crossing. Comparing the result of open-pollination with that from crossing, apart from one case, the depression varied from 2 % to 26 %. There was also some inbreeding effect on bole straightness and when this effect was added to that on height (as represented by relative dominance), scores for progeny from open-pollination were depressed in relation to those from cross-pollination by 8 % to 13% in three tests at age three years. Bees were not often seen on flowers which had been open for more than two days, and this fact is seen as a natural barrier to selfing, in that when they are most receptive to self pollen, stigmas are not often pollinated by bees. In addition, pollen is easily lost from unreceptive stigmas and this, combined with a considerable degree of protandry, is seen as controlling the amount of pollen remaining on the stigma for possible germination later. But there was an overriding barrier to selfing, in that there were invariably fewer self-fertilised seeds than cross fertilised seeds after hand pollination and there were strong indications that when mixed- (self- and cross)-pollination was done, selective fertilisation occurred in favour of the cross pollen. It is necessary not only to know about such features of the reproductive system, but also to take appropriate action. Some self-fertilised seedlings are liable to be present in the natural progeny and these are subject to inbreeding depression in height and form. Some of these seedlings may be removed by culling in the nursery, but it would still be necessary to do testing and rejection of the more self-fertile clones. Hand-pollination in the seed orchard is another possible way of reducing selfing and this would be at least theoretically effective.