Reproductive Biology of Pohutukawa (Metrosideros Excelsa) (Myrtaceae)

Abstract

The objective of this project is to describe and analyse the reproductive biology of pohutukawa, by integrating information from floral biology, breeding system and pollination biology. New Zealand pohutukawa (Metrosideros excelsa Sol. ex Gaertn). a member of the Myrtaceae, is a mass-flowering tree endemic to northern New Zealand coastlines. Compound inflorescences develop over a period of ten weeks in six morphologically distinct stages. Trees flower over a peak period of two weeks, and the inflorescences contain an average of 14.3 large, hermaphrodite, red 'brush' flowers that remain open for seven days. Most pollen is viable (93.6%) and the receptivity of the wet-papillate stigma extends for at least nine days. Each flower produces approximately 46 μL nectar per day, containing l8% (w/v) sucrose. Neither dichogamy nor herkogamy prevent pollen and stigma interference, and floral design and display are consistent with high levels of autogamous and geitonogamous self-pollination. The stigmatic exudate of unpollinated pistils stains intensely for carbohydrates, lipids and proteins, but shows a notable decrease in lipids and proteins following pollination. The style has a solid transmitting tissue with large mucilage-filled intercellular spaces which stain weakly for polyanions and pectins irrespective of pollination. Although starch grains in the stylar cortex are depleted following pollination, stylar resources in general appear to be sparse and this results in a low speed of pollen tubes (2 mm / d) through the intercellular spaces of the transmitting tissue. Seed capsules of pohutukawa contain a mixture of fertile (embryo-containing) and infertile (embryo-lacking) seeds. Fertile seeds weigh approximately 0.15 mg. stain positively with 1% tetrazolium chloride, and are randomly disposed on the placenta. Their germination rate exceeds 90% after up to one year of cold storage, but decreases rapidly when stored at room temperature. Controlled pollinations with self- and cross-pollen from single donors and a pollen mixture from five unrelated parents showed that seven out of ten trees were self-incompatible, suggesting that natural populations may consist of a mosaic of self-incompatible and self-compatible individuals. Self-incompatibility is late-acting as pollen tubes from selfs and crosses reach the ovary simultaneously l0 - 15 d after pollination. In common with other Myrtaceae, the seed / ovule ratio in pohutukawa is low and not limited by the stigmatic pollen load. The pollen / ovule ratio of 462.5 (SE ±43.4) places the breeding system of pohutukawa between facultative selfing and facultative outcrossing. Mainland populations of pohutukawa have been reduced to fragmented stands, and the original suite of bird pollinators has been largely replaced by introduced species. In contrast, the native pollinator fauna of several offshore islands remains intact, including the three species of the New Zealand honey eaters (Meliphagidae) and solitary bees. Using allozyme analyses. multilocus outcrossing rates were estimated for Little Barrier Island and Tiritiri Matangi Island and for three mainland populations in comparison. They were among the lowest in the Myrtaceae (tm = 0.22 - 0.53) and the loss of native pollinators has no measurable effect on the mating system. Although there is no difference in the germination percentage of fertile seeds from self- and cross-pollination treatments (98.4%), 'selfed' seedlings show marked inbreeding depression in height after six months. Wright's fixation index is consistently higher for seedlings (Fs) than for mothers (Fm) in all populations, indicating that selection may eliminate selfed offspring from populations prior to reproductive maturity. Exclusion experiments were undertaken on Little Barrier Island to assess the effect of native birds and bees on outcrossing and seed production. In bird exclusion experiments in the lower canopy (2 - 4 m) with flower access to bees only, estimated outcrossing rates were lower (tm = 0.40) than in open pollination (tm = 0.58), suggesting that bees effect more self-pollination than birds. The highest outcrossing rates (tm = 0.71) were found for open pollination in the upper canopy (> 4 m). Numbers of fertile seeds per capsule were 45% higher after open pollination than in treatments with bee visitation only, and 28% higher than in treatments where all flower visitors were excluded. The results suggest that native bees visiting pohutukawa flowers reduce seed set and effect less outcrossing than birds, and that a large proportion of seeds arises from automatic self-pollination. In trees of a modified mainland population with predominantly introduced birds and a mixture of introduced and native bees there was no decrease in seed set for the treatment allowing flower access by bees only, indicating that - in contrast to native bees - honeybees did not reduce seed set in pohutukawa. In conclusion, although the floral biology of pohutukawa permits geitonogamy, a combination of outcrossing (predominantly by bird pollinators), self-incompatibility, and inbreeding depression act to maintain heterozygosity and result in the production of sufficient offspring that will ensure the survival of the species.