A project undertaken at Kings Park Science, Department of Biodiversity, and supervised by Dr Siegy Krauss
Although most plants are insect pollinated, more than 900 bird species pollinate members of ca. 500 of 13,500 vascular plant genera. Australia has over 110 bird species (mostly honeyeaters and lorikeets) that have been recorded visiting the flowers of >1000 plant species from > 64 genera, with many of these genera (Banksia, Grevillea, Eucalyptus, Adenanthos, Anigozanthos) highly diverse and abundant. These birds are often highly mobile generalists and locally abundant due to copious nectar production, leading to frequent aggressive interactions within and among species. The behaviour of these birds interrupts hypothetically optimal foraging behaviour. As such, the foraging behaviour of these birds suggests greater pollen carryover from more pollen donors compared with insects. This is predicted to increase the ability of plants to avoid selfing and mate with multiple other partners, resulting in genetic benefits for the offspring of plants they pollinate.
Figure 1.
We tested these predictions for banksias in South Western Australia (SWA). SWA is not only an internationally recognised biodiversity hotspot, but an international hotspot for pollination by vertebrates. Here, 15% of its ca. 8000 plant species, and 40% of threatened flora, are pollinated by nectarivorous birds and marsupials. As such, the genetic consequences of vertebrate pollination are critical for understanding both the evolution of the flora and how best to conserve it.
Selective pollinator exclusion experiments were employed to assess the impacts of different classes of visitors (birds, mammals, insects) to the flowers of banksias. These impacts were assessed in terms of seed quantity, as well as seed quality through a genetic determination of mating system parameters and paternal diversity, followed by survival and growth of seedlings in field trials.
We have found:
(i) for the tree B. menziesii, birds, honey possums and insects all contributed to pollination in varying degrees, but birds were the most effective pollinators because their foraging behaviour promotes full outcrossing and high paternal diversity. In contrast, when birds were excluded from flowers, honey possums and insects (primarily European honeybees) promoted self pollination, and there was strong selection against the products of selfing (e.g. higher early mortality of selfed seedlings than outcrossed seedlings). This fitness advantage for plants from nectar-feeding bird pollination suggests a mechanism to explain its abundance in SWWA.
(ii) for the shrub B. catoglypta, which posseses floral traits suggesting adaptation to pollination by mammals, similar experiments resulted in little difference in seed quantity and quality, suggesting birds, mammals and insects (primarily introduced honeybees) are largely equivalent in their pollinator effectiveness. These results do not support the hypothesis that the flowering traits of B. catoglypta are adaptations specifically to mammal pollination. Nor do they offer strong support for a benefit from bird pollination. Rather, these plants possess traits that encourage generalist pollination and maintains complete outcrossing through self-incompatibility.
(iii) extension of these studies to four other shrub Banksia species of the sandplains north of Perth, with contrasting floral traits further challenged the assumption that in Banksia floral crypsis and strong scents are adaptations to primary pollination by honey possums. Instead, they suggested a spectrum of vertebrate and insect pollination systems, with rodents identified as key pollinators of some species.
(iv) insight into surprisingly low floral visitation by honey possums of the tree B. attenuata and a potential role for nocturnal pollination by beetles for the summer flowering species, with identification of novel compounds in floral scent that serve to attract beetles. This novel finding adds to the known diurnal pollinators (birds, insects) and nocturnal mammals, and extends the generalist-pollination syndrome for banksias to include, in some cases, beetles. It also highlighted the neglected side of nocturnal pollination.
Ultimately, by extending the traditional assessment of pollinator effects on plants beyond seed set, to include an assessment of genetic and offspring consequences, we have demonstrated that variation in pollinator effectiveness extends beyond pollen deposition and fruit set mechanisms to also include the genetic quality of pollen deposited. Our results have shown how a genetic component of pollination that increases offspring fitness could favour the evolution of bird pollination. We have also confirmed the generalist pollination syndrome for many of these banksias, and highlighted a need for further research into nocturnal pollination