Two native mistletoes need tuis and bellbirds to open their specialised "explosive" flowers. The large red flowers spring open only when tweaked by these birds.
The two endemic Peraxilla species both have large red showy flowers, unlike most New Zealand plant species which have small and inconspicuous flowers. We discovered in 1993 that both Peraxilla species have specialized "explosive" flowers that require birds to tweak the buds before they will open (see Ladley and Kelly 1995). These explosive flowers have been reported in other Loranthaceae mistletoes, especially in Africa and the Near East. However, the Peraxilla species are the first explosive mistletoe flowers reported from Australasia.
To open flowers, birds grasp the top of the bud with their beaks and twist. This causes the flower petals to spring open (in less than ¼ of a second), and the bird can then insert its beak to drink nectar and thereby pollinate the flower. Flower buds that are protected from all potential pollinators do not open their petals, although the petals do eventually separate as a unit from the base of the flower.
The main birds that can twist open Peraxilla flowers are the tui (Prosthemadera novae-seelandiae) and bellbird (Anthornis melanura), which are both endemic honeyeaters (Meliphagidae). A few non-endemic birds such as native silvereyes (Zosteropis lateralis) and exotic chaffinches (Fringilla coelebs) occasionally open flowers but this is uncommon.
However, we discovered in 1996 that native insects (solitary bees) also sometimes open Peraxilla flowers. Native bees are able to bite the top of ripe P. tetrapetala buds (but not the larger P. colensoi buds), which causes the flower to spring open and enables the bee to collect pollen (their principal larval food supply) and nectar (see Kelly et al. 1996). This is the only known example of an invertebrate that opens an explosive, vertebrate-adapted flower. (Click here to watch our video footage of a Leioproctus opening a P. tetrapetala bud at Lake Ohau).
Two native bee species open P. tetrapetala flowers: Hylaeus agilis (Colletidae) and another native Colletid bee, Leioproctus sp "tarangahape" (taxonomic revision of this genus is currently underway). These bees are particularly important in areas such as Lake Ohau, where there are now few native birds that can open mistletoe flowers. This area supports many bees, so little pollen is left in the flowers that have already been opened. This makes it advantageous for bees to be able to open new flowers so that they can harvest more pollen.
In some places, there are not enough pollinating birds to open all of the mistletoe flowers, presumably because of predation by introduced mammalian predators such as cats and stoats. While the flowers can set some seeds by self-pollination even in an unopened bud, the level of seed set is much lower (c 2-5%) than with good bird pollination (35-60%). (See Robertson et al. 1999 describing how mistletoes are pollen limited on the mainland of New Zealand).
Another experiment showed that at one site with pollination limitation (Craigieburn), mistletoe fruit and nectar were preferred food sources when they were in season. Bellbirds ate P. tetrapetala nectar and fruit more often than would be predicted based on their relative contributions to total available food at this site (Murphy and Kelly 2003). On the other hand, bellbird counts were low, even compared to other sites in the eastern South Island, which in turn have lower counts of bellbirds than the western South Island and offshore islands. This suggests that the low number of bellbirds at Craigieburn, rather than their choice of diet, is responsible for the pollination limitation at this site, and by implication also at other South Island sites (see Murphy and Kelly 2001).
Native honeyeaters have declined on the New Zealand mainland because of a decrease in available habitat, and more importantly, because of introduced predators such as stoats, cats and rats. At sites where these predators are controlled, honeyeater numbers increase dramatically. In 2001-2202, we conducted an experiment to test whether controlling stoats (Mustela erminea) at Craigieburn might also indirectly benefit mistletoe reproduction, by increasing honeyeaters that pollinate the plants.
A total of 33 stoats were trapped in the treatment area during the summers of 2001 and 2002. This significantly reduced stoat numbers in the area compared to a nearby non-treatment area. This decrease in stoats had significant benefits for bellbirds at Craigieburn. After trapping, fledging increased by 10-fold, and bellbird density increased by 80%. Unfortunately, however, no effect of this increase in pollinators could be detected for mistletoe reproduction. Pollinator visitation and fruit set did not increase with higher bird densities. The reasons for this are not clear, but current work is exploring plant/bird interactions in places with very high bird densities (such as the Rotoiti Nature Recovery Project in Nelson Lakes National Park) to try and understand how bird densities interact with mistletoe flowering density to determine pollination rate.
Recently, the team has begun to look at other bird-pollinated plants in New Zealand to determine whether pollination limitation might affect other species besides mistletoes. So far, we have found evidence that Fuchsia perscandens may also be suffering from low honeyeater pollination (see Montgomery et al. 2001). We have found that fruit set of hand-pollinated flowers was at least 1.7 times higher than un-manipulated flowers, and fruit set of un-manipulated flowers did not differ significantly from that of bagged flowers where all pollinators were excluded. The study sites on the Port Hills (Christchurch) have low densities of honeyeater birds, and the birds may be reluctant to forage near the ground where the F. perscandens flowers are.
We are also currently carrying out experiments on other bird-pollinated species like tree fuchsia (Fuchsia excorticata; see Sessions 2001), kowhai (Sophora microphylla and Sophora prostrata) and Rhabdothamnus solandri to see how widespread bird-pollination failure might be.
Another question that our research has investigated is how fragmentation might affect mistletoe pollination. Studies at Lake Ohau, in the central South Island, suggest that mistletoes do well in forest fragments with lots of light around the edges. We found that mistletoe density was 2-3 times higher on fragment edges than in the forest interior. Flower predation by a native caterpillar (Zelleria maculata) also decreased on edges from 48% of flowers in continuous forest to 8% on isolated trees.
In addition, the mistletoes on the edges of forest fragments get better pollination service from birds than mistletoes in forest interiors. Fruit set increased 4.4 fold on edges compared to forest interiors. Pollination was lowest in continuous forest (14% seed set) and highest on isolated trees. We have also shown an increase in bee visitation on forest edges at Ohau and at a second site (Craigieburn). Bees may need sunshine to keep warm at sites that are frequently cold and windy. All of these data suggest that a limited degree of fragmentation improves Peraxilla reproduction by enhancing pollination, as long as enough forest habitat survives to maintain bird populations (see Kelly et al. 2000).
Since 1992, a team led by Dave Kelly, Alastair Robertson and Jenny Ladley has been investigating the reproductive biology of six New Zealand mistletoe species in the family Loranthaceae. The research covers pollination, seed set, dispersal, germination and establishment of seedlings, but the area of greatest interest so far has been pollination.