Further notes to Summary Points
Point 1
The geological history of the islands suggests that a significant – but variable – area of land has existed at Norfolk Island for more than 2 ma. Changes in sea level associated with fluctuations in global ice volume meant that, at the lowest sea stand of c 120 m below present levels (Fairbanks 1989) about 18,000 years b.p., the sub aerial area was c 320,000 ha, as measured from the local bathymetry (Main & McKnight 1981). Rising sea levels reduced that by nearly two orders of magnitude in a few thousand years or less. The rapid flooding of the lowland areas surrounding the present Norfolk and Phillip Islands isolated plants and animals on these two remnants: far from having evolved independently on each island, it is likely that the pattern of endemism within the group reflects the former distribution of organisms and their chance survival on the remaining high points. Extinctions probably resulted, but their extent is unknown and unknowable. However, given the cyclic changes in sea levels over the past 2 Ma the fauna and flora of greater Norfolk Island are likely to have undergone parallel cycles of recruitment, expansion, contraction, and local extinction with the waxing and waning of the sub aerial area. Such changes are similar to those that have been described for various island systems in the Caribbean, as, for example, Bermuda (Olson et al. 2005).
However, the present reduction in area of natural vegetation, and associated extinctions of both plants and animals, is probably unprecedented in the history of the islands. Many of the vertebrate extinctions were of burrow-nesting and surface-nesting petrels and shearwaters. Holdaway & Anderson (2001) discussed the former diversity of these birds in the Norfolk Island group; six species are known to have bred in the island group and another two or three may well have, before Polynesians arrived and brought the Pacific rat Rattus exulans.
Petrels and shearwaters are relatively old geologically, with the genus Puffinus (shearwaters) being known from at least the mid Miocene with “very little change within these lineages in 15 million years or so” (Olson 1985: 211). Hence, it is likely that as soon as the rocks of proto-Norfolk and proto-Phillip Island had cooled sufficiently, sea birds very similar to those present in the recent past colonised and multiplied into large populations, occupying most areas of land. The large populations of petrels breeding today on Macauley Island (Kermadecs) and in the historic past on Raoul Island (also in the Kermadecs), after geologically very recent (<< 10,000 years ago) “sterilisations” by violent volcanic eruptions, demonstrate that these and other birds can build up to large numbers on new islands very quickly. Until they were decimated by human interventions, there is no doubt that petrels, shearwaters, and other seabirds were abundant on Norfolk and its outlying islands throughout their post-eruptive history.
Point 2
Some species of petrel were probably eliminated before European settlement in 1788. Pycroft’s petrel (Pterodroma pycrofti) was commonly represented in a natural deposit excavated by Robert Varman (Holdaway & Varman unpubl. data), and a significant component of the food remains left by the early Polynesians at Emily Bay (Holdaway & Anderson 2001). It may have been present in small numbers in 1788 (Holdaway & Anderson 2001) but was certainly extinct soon afterwards. The Providence petrel Pterodroma solandri was famously extirpated from the island in the 1790s by the European settlers after HMS Sirius foundered off Slaughter Bay (Hoare 1987). Nearly 200,000 birds were taken in a single season from the area of Mt Pitt, but the total over several years may well have reached 1 million.
After the First Settlement was abandoned, a population of petrels or shearwaters was re-established in the Mt Pitt – Mt Bates area. The first evidence for this recolonisation was found in the stable isotopic composition of a large fallen Norfolk Island pine tree beside the Red Road Track, but eye-witness accounts have been obtained of the presence of such birds during World War II (I. Baumgart, letter to M. Christian 2005). Their extinction during that period can probably be attributed to disturbance and direct predation by the armed services, as a result of the birds interfering with the performance of the radar station on Mt Bates and the field emplacement on Mt Pitt. With the removal of these birds, the only marine nutrient source on the higher ground of Norfolk Island itself was the scattered pairs of tree-nesting white terns Gygis alba., which in their present numbers would contribute an insignificant amount to the forest nutrient flux.
At present, the only burrow-nesting seabird still breeding on Norfolk Island in numbers is the wedge-tailed or Pacific shearwater Puffinus pacificus. There are no recent estimates of numbers, but the size of “rafts’ of this species – gatherings of birds off the coast before they venture ashore to their burrows – has apparently decreased over the past few decades, suggesting an ongoing population decline as a result of predation at the nest site. Occupied burrows are still present at the edge of many coastal cliffs, and where predators are controlled, even occur inland on flat ground. This habit may have contributed to the extensive areas of enhanced nutrients in the Selwyn and Steels Point soils on level areas around the island.
Small numbers of Norfolk Island little shearwaters Puffinus assimilis may still attempt to nest on Norfolk Island; cats destroy most of the black-winged petrels Pterodroma nigripennis that alight on the island. Black-winged petrels are unknown in fossil and historic records from the island as a breeding species, but it is presently a vigorous coloniser from its main breeding ground at Macauley Island in the Kermadecs.
Point 3
Given the high transfer rates estimated for burrow-nesting petrels elsewhere (up to 1 tonne ha-1 yr-1 for the 250 g Hutton’s shearwater Puffinus huttoni in the Seaward Kaikoura mountains of the South Island of New Zealand), the population of Providence petrels (Pterodroma solandri) alone, estimated to be > 100,000 pairs and likely to have been at least 500,000 pairs, would have brought in 60 – 300 tonnes of high nutrient guano alone (without eggs, dead birds, and other sources of oceanic nutrients). If the present, almost certainly declining, population of wedge-tailed shearwaters breeding along the coastal cliffs is of the order of 5000 pairs, it may be contributing at least 3 tonnes of fertiliser to the Selwyn and Steels Point soils and cliff faces each year. At the present cost of triple superphosphate (without N supplement), the present annual input would be worth c. NZ$2500; the input to the Mt Pitt area 220 years ago would have been equivalent to fertiliser worth a minimum of NZ$45000 up to NZ$225000 per annum.
Bancroft et al. (2005) reported that soils in a wedge-tailed shearwater colony on Rottnest Island had 470% higher levels of nitrate, 118% higher of phosphorus, 102% of ammonium, 69% of sulphur, and 34% of potassium than in heath soils around the colony. The shearwaters deposited guano at over 230 kg (made up of 50.9 kg N, 5.7 kg K, 5.5 kg S, 3.6 kg P) ha-1 yr-1 (dry mass).
The soil analyses done by Stephens & Hutton (1954) demonstrate that the amounts of P in the surface soil in most of the island is well below that of the parent material, because of the limited weathering of the parent material now. Those soils where seabirds still breed, or did so within the past 50-60 years, had levels of P much higher than could have come from the parent material, even when freshly weathered.
A potential source of both P and other minerals, and of the thick soil and weathered layer characteristic of Norfolk Island, is aeolian dust from Australia. The annual rate of dust accumulation in the South Pacific is estimated at about 0.35 g m-2 yr-1, which equates to about 0.45 mm yr-1. To develop a layer up to 45 m or more thick over the 2.3 ma life of the island would take a deposition rate of about 10 mm yr-1. The deposition rate may have been substantially higher during glacial periods, when central Australia was even more arid than today, but even then it is unlikely that the amount of dust arriving on Norfolk Island could account for the levels of P at present in some soils.
Stephens & Hutton (1954) and Hutton & Stephens (1956) made some interesting general observations on the pH and amounts of exchangeable calcium (Ca2+) and magnesium (Mg2+). They point out that the usual excess of rainfall over evapo-transpiration must leach the soils, which should result in low pH soils low in exchangeable metal ions “unless some source has supplied calcium and magnesium”, and that the proportions of these elements probably ruled out aerosol sources. Instead, they invoked the former presence of a raised coral limestone “makatea”. The soils enriched in exchangeable Ca2+ and Mg2+ are at the periphery of the island, which would be where such a makatea structure would have been. Radiometric dating has shown that the volcanism that formed the islands occurred very much later than the early Tertiary or older age suggested by Hutton & Stephens (1956), but being at least 2.3 ma old, the rocks have been subjected to levels of weathering that make them poor present sources of plant nutrients (Vitousek 2004). Ca2+ is at a high concentration in guano (Bancroft et al. 2005) and the relatively low annual rainfall (c 1350 mm) would maximise its retention in the soil.
From the data presented by Stephens & Hutton (1954) it seems likely that the chemistry of the Norfolk Island soils reflects in large part the past distribution of breeding sea birds. The levels of P in the Selwyn and Steel’s Point clays on such a deep weathering blanket, are likely to be derived from burrow-nesting sea birds. As noted above, the Palm Glen clay also shows evidence for residual P, probably from birds on the slopes above, whose “skeletal” soils were not analysed. The low levels in the Mt Pitt clays probably reflect their position farther down slope, and the transition to the relatively low P soils of the main plateau.
Point 4
Croll et al. (2005) showed that the extinction of seabird colonies on many of the Aleutian islands after the introduction of arctic foxes (Alopex lagopus) reduced the transfer of marine nutrients to the terrestrial ecosystems to the extent that fertile grasslands changed to infertile shrub tundra. The authors do not make the point, but the changes in vegetation were so great that if they had been recorded in the pollen record without information on the predator or the seabird populations, they would have been attributed to a sudden climate cooling event. Fertilisation of plots on islands with foxes resulted in a return of the grasses, showing that it was the nutrients brought in by the seabirds that had maintained the terrestrial ecosystem.
There are high concentrations of plant-available P in seabird colony soils (Bancroft et al. 2005; Croll et al. 2005; Hawke & Holdaway 2005).
Point 5
Petrels and shearwaters are landscape architects, and move substantial volumes of soil during burrow making and repair. Bancroft et al. (2005) showed that wedge-tailed shearwaters burrowing on Rottnest Island significantly altered the physical and chemical factors in the soil, such as bulk density and hydraulic conductivity. They reported that colony soils had 470% higher levels of nitrate, 118% higher of phosphorus, 102% of ammonium, 69% of sulphur, and 34% of potassium than in heath soils around the colony. The shearwaters deposited guano at over 230 kg (made up of 50.9 kg N, 5.7 kg K, 5.5 kg S, 3.6 kg P) ha-1 yr-1 (dry mass). Westland petrels breeding on the West Coast of the South Island, New Zealand contribute about 38% of the nitrogen in freshwater stream insect larvae more than 1 km below the colony (Harding et al. 2004), and in the herbivorous New Zealand pigeon (Hemiphaga novaeseelandiae) – a close relative of the extinct Norfolk Island pigeon, H. spadicea (Hawke & Holdaway 2005).
Point 6
Because the past two centuries have been the only time in Norfolk Island’s subaerial history when burrow-nesting seabirds have not been a significant component of its biota, it is likely that the flora of the island has evolved in the presence of significant inputs of nutrients from the oceanic ecosystem and that the loss of that input will have changed their growing conditions. The evidence from the geology and soils suggests that there is little input of the essential nutrients from rock weathering, so the natural vegetation within the Norfolk Island section of the Park is now subjected to unnaturally low levels of essential nutrients. The soil fauna and nutrient cycles are now almost certainly also significantly different, with even leaf retention being affected by changes in nutrient regimes (Vitousek 2004), leaves now being retained longer, so less litter fall, so that nutrient cycles will be much slower than in the past. It follows that restoration of burrow-nesting petrels will be essential to the survival of the Norfolk Island terrestrial ecosystems.
Removal of the feral chickens (Gallus gallus), which scratch up the forest soil and consume significant quantities of invertebrates associated with litter break down and releasing nutrients, is essential.
Point 7
Evidence from the North and South Islands of New Zealand (Holdaway 1999) demonstrates that small to medium-sized burrow-nesting petrels, and particularly species in the genus Pterodroma, cannot coexist with mammalian predators in the breeding colonies. This is also shown emphatically by the inability of black-winged petrels to permanently colonise the periphery of Norfolk Island in the presence of predation by cats; only where cats and rodents are controlled has this species gained a foothold. Holdaway & Anderson (2001) demonstrated that the small species of Pterodroma that formerly bred on Norfolk Island (Pt. pycrofti) was almost eliminated by predation by Pacific rats in prehistory, with very few surviving into the European period.
Point 8
It follows from Point 7, and the continued inability of small petrels to maintain viable breeding populations on Norfolk Island, that the predators must be removed and not merely monitored. Small populations of rats can keep islands off New Zealand essentially free of breeding petrels; the best (worst?) example is that of Campbell Island, where Norway rats (Rattus norvegicus) and cats eliminated from the main island all of the smaller petrels – and most of the larger species, and the endemic species of teal and snipe, and local race of the New Zealand pipit (Taylor 2000). Pyrcroft’s petrel is recovering its numbers rapidly on the islands in the Hen & Chickens and other island groups off northern New Zealand from which the Pacific rat (Rattus exulans) has been eradicated.
Point 9
Experiments and pilot programmes have demonstrated a capacity to establish viable populations of small burrow-nesting seabirds in areas formerly inhabited by the species. Examples include the translocation of fairy prions and common diving petrels to Maud Island in Cook Strait, New Zealand (e.g., Miskelly et al. 2004), and of fluttering shearwaters to Maud Island, Marlborough Sounds, New Zealand (Bell et al. 2004). Efforts are under way to return breeding populations of Pterodroma petrels to sites on the North Island, New Zealand, with predator-fencing and predator removal. Research is under way to establish parameters such as minimum weights for translocated chicks, and optimum feeding effort and weight gains for translocated chicks.
Point 10
The only conclusion that can be drawn from the information available now is that if predators, including all three species of rodent, and cats, are not removed from Norfolk Island, or at least initially a significant part of the Norfolk Island National Park, then the natural processes and systems which supported the Norfolk Island terrestrial ecosystem will continue to be degraded, endangering both the natural vegetation and fauna (invertebrates and vertebrates) in ways that cannot be prevented by present policies and practices of monitoring and control. The previously unrecognised former contribution of seabirds to the island’s terrestrial systems was pivotal. Without the restoration of the “free” nutrients from the surrounding ocean, Norfolk Island will become another fully degraded Pacific ecosystem. If they can be reinstated, and that can happen only after mammalian predators are removed from significant sections of the island, and preferably the whole island, then the restored ecosystem will very soon become the symbol of environmental restoration in the Pacific and around the world. The island’s economy will benefit accordingly.