Location and Geology of Nikumaroro

Jump to: navigation, search

Nikumaroro: Physical Character

Howard Alldred’s Notes.

Need to locate Howard's original illustrations, if possible.--MXM.


Nikumaroro, formerly known as Gardner Island, is a typical Central Pacific coral atoll. It is located just south of the equator, due north of Western Samoa and the Tokelau Islands, at 04º 30' S, 174º 30' W (Fig. 1). The atoll is 7 km long by 2.5 km wide, elongated northwest-southeast, having grown on the crest of a submarine volcanic cone with a similar orientation that rises out of 5000m of water. <embed src='I23jungleDM.jpg>

Fig. 1: Nikumaroro is an island in the Phoenix Group. It is located just south of the equator, in a line due north from Samoa, via the Tokelau Islands.

The island forms part of a volcanic chain that includes Carondelet Reef to the south and a substantial seamount to the north with dimensions similar to Nikumaroro’s. Other islands in the Phoenix group, such as Orona, Manra, McKean, and Kanton, lie on parallel volcanic chains.

The atoll is made up of a central lagoon, a surrounding terrestrial rim, an extensive ocean reef flat, and a steep drop-off into the ocean abyss (Fig. 2).

Fig. 2: Nikumaroro is an elongated atoll, 7km long and 2.5km wide, formed on the crest of a submarine volcano, rising through 5000m of water. It has a central lagoon served by two passages on the leeward (southeastern) side.

The lagoon is relatively shallow, on average no more than 5 metres deep, and is studded with coral heads and patch reefs. Most of the lagoon shore lacks beach; a few short stretches of sand arc around the northwest end. Commonly, heavily vegetated coral shelves drop directly into the lagoon, or into a swampy foreshore of calcareous ooze, a mixture of bird droppings and coral erosion products. Water clarity varies greatly depending on weather conditions. When turbid it is virtually opaque; when calm it can be fairly clear. It is usually a bright aquamarine-turquoise in color.

The terrestrial rim is built on Holocene, and perhaps Pleistocene, beachrock ramparts that have provided a solid foundation for the overlying sediment. This sediment mantle has been built up by deposition from multiple overwash events that happen during severe storms. A mineral exploration survey report in 1978 described the island as "a series of cemented coral rubble platforms."

The rim is continuous along the windward, northeastern side, but the continuity of the southwestern, leeward side is interrupted by two passages. The northern passage – Tatiman Passage – is a semi-navigable channel, while the southern passage – Baureke – is a usually non-navigable overflow/overwash channel.

On both sides of the island, though more noticeably to windward, there are ridges or berms of coral rubble just behind the beach in many places, the results of storm surges. The highest point of the terrestrial rim, near the northwest end, is elevated only about 6-metres above sea level.

The prevailing wind is from the east-northeast, as is the set of the swells (Figure 3). The windward (northeast) shore is a long sandy beach about 20 to 40 meters wide, varying with the tide, behind its. The southwest side of the island, normally in the lee, takes the brunt of periodic serious storms from the southwest and NW. Its shoreline is variable both in width and character. In some places shelves of coral, often heavily overgrown, drop directly to the reef flat, while in other places there are beaches, but these are for the most part narrower, steeper, and rockier than on the windward side, typically made up of finger-sized coral rubble mixed with sand. There are substantial stretches of sandy beach toward the northwest end, however.

Outboard of the terrestrial rim is an extensive reef flat, varying from 200 to 250 metres wide. From the toe of the beach, the reef surface comprises the typical geomorphic succession of features (Fig. 3). The inner section consists partly of discontinuous ridges of beachrock, which presumably date from the early Holocene (5000 - 10,000 y.B.P.).

Fig. 3: Generalised geomorphology of Nikumaroro, showing the principal reef, rim and lagoon features.

Either in association with the beachrock outcrops, or as a result of its absence, a “boat channel ” runs along the inboard edge of the reef flat. Where it is actually present, the boat channel is often no more than a series of discontinuous channels, troughs and pools (Fig. 4). The walls of these pools are usually steep-sided, and they are commonly less than 1 metre deep. This zone is on average 130m wide.

Fig. 4: Photo of the “boat channel”, which on Nikumaroro, where it is actually present, is a non-navigable series of discontinuous channels, troughs and pools, not usually more than 1-metre deep. They form temporary sediment traps.

Seaward of the boat channel is a more-or-less flat coralline pavement that often has a pocked surface and is criss-crossed with fractures and tension joints. The surface is slippery with the growth of the coralline algae Porolithon, which must be washed on a daily basis by seawater to maintain it. The pavement averages 70m in width.

The outer edge of the reef comprises the “spur and groove” zone (Fig. 5). The groove consists of a steep walled, narrow canyon, typically several metres deep, that starts at the reef edge, and extends in a direction perpendicular to the reef edge, back towards the shore as a gradually shallowing and tapering slot. Grooves form in the high-energy surf zone and act as gutters to channel water draining from the reef surface. Grooves are commonly spaced at 20 to 50 metre intervals and can stretch back into the reef pavement up to 50m.

Fig. 5: Photo of a groove tapering back into the pavement surface from the reef edge. They act as water chutes funnelling water up onto the reef, as well as drainage channel for water returning from the intervening spur surface back to the ocean.

The intervening spurs are tabular to slightly convex rather than rounded as the name implies, but have sufficient profile to allow water to be shed from the reef surface. The swash is channelled back out to the ocean via the grooves. Like the reef flat, the spur surfaces and even the groove walls are covered with slippery Porolithon algae.

Large boulders are present on the reef surface in the spur and groove zone. These have been broken off and thrown up by the hydraulic hammering of surges underneath reef edge overhangs.

Beyond the reef edge is the steep drop-off zone. The average gradient of the upper submarine slope down into the abyss is about 40º. However, the surface is far from smooth, with ledges and precipices stepping down that reflect Pleistocene sea level still-stands. There is usually a prominent submarine terrace at -50m, but others occur at -10 and -15m.

At the nadir of the last glaciation, sea level stood at -130m. Below this depth, the flanks of the volcano are likely to be covered with a mantle of tumbled material, from sand and gravel to boulders and blocks. The flanks of the volcano are not smooth surfaces but are fluted, with wide-spaced ridges and deep valleys. Transport of sediment into the abyss is down the floors of these valleys.

The atoll is subject to semi-diurnal (equal-amplitude, twice daily) tides. There is also a marked spring-neap tidal dichotomy.


In 1937, Harry Maude commented that "the profuse vegetation on Gardner Island gave it the appearance, from the sea, of possessing several low hills". The same can be said today. Nikumaroro retains significant stands of the indigenous tree called “buka” in I Kiribati (Pisonia grandis). Bukas are impressive trees, up to 20 meters high with a canopy extending in a radius up to 8 meters or so from the thick, gray-barked trunk. Major buka stands are in the northwest, with a fringe running down the narrow land behind the windward shore. Substantial buka stands have been cleared behind the lee shore and on the southeast end (compare Figures N-3 and N-4 [Based on airphotos, show sequence of clearing]).

According to Maude, in 1937 the island supported "some fine groves of 'kanawa' trees (Cordia subcordata). The fine-grained kanawa wood was much sought after for the construction of furniture and containers; it will feature importantly in subsequent chapters. A major grove occupied the area known as "kanawa point", and others may have grown on the northeast side of the lagoon on the land called Taraea. Kanawa seems to have been completely cleared from the island, neither we nor a botanical survey party that visited the island in 1978 noted any specimens.

Bukas are today being seriously crowded by introduced coconut palms. Even where the bukas have not been cleared to make way for coconut plantations, feral coconuts have grown up along the fringes of the buka stands and compete with the indigenous trees for light and nutrient. “Te mao,” (Scaevola sp.) is a major competitor as well, and a major impediment to human travel around the island. Mao presents a tangled jungle of interwoven stalks, each up to about 5 cm. in diameter, making up interwoven bushes up to 3 meters high. When fresh, mao is easily cut with a bush knife, but when dry it defies non-mechanized human attack. Since Nikumaroro experiences periodic dry conditions, most mao bushes are made up of dry stalks masked by a deceptive “skin” of green ones. mao can be virtually impossible to cut through, yielding only to a chain saw.

Mao tends to be concentrated along the ocean and lagoon shores, but can extend 50 or more meters inland, so on narrower parts of the island its coverage is virtually continuous. [[RenTe ren, -- Tournefortia argentea, is also common along the shores, though far more scattered than mao. Te ren is a small tree, on Nikumaroro growing to a height of perhaps six or seven meters.

Other plants recorded on the island include a small tree called "non" (Morinda citrifolia), a bush known as kaura (Sida fallax) and a creeper called boi (Portulaca sp.).


The only mammals on the island today (as far as we know) are rats, which are quite numerous. Feral dogs and cats have been reported; we observed one not-too-long dead cat corpse in 1989, and occasionally have heard what sound like cat cries in the forest, but have never seen a live representative. Dogs were subjected to a deliberate eradication program in the 1980s, which seems to have been successful. Indigenous terrestrial animals fall into three roughly defined practical categories: crabs, birds, and insects/arachnids. Crab varieties include Coconut or Robber Crabs (Birgus latro), other hermit crabs (Coenobitidae), and smaller shore crabs (Gelasimus). Large numbers of very small hermit crabs certainly represent juvenile stages of Birgus latro, but there are many larger hermit crabs as well. Common bird species include snowy tern (kiakia in I Kiribati), Frigate Birds (Boobie, Gannet), and Red-Tailed Tropic Birds. Insects and arachnids include ants, bees, and many web-spinning spiders. A single scorpion was noted in 1997.

Offshore the fauna is much more extensive and varied. Fish include Grouper, Trevally, Barracuda, Blacktip Sharks and Whitetip Sharks, Gray Reef Sharks, and a wide variety of other reef and deepwater species. Sea turtles are common, and come ashore in some numbers along the windward side to lay eggs. Small reef sharks are common in the lagoon.

Because coconut crabs may have played roles in the fate of Earhart and Noonan, some detail about their character and behavior is in order. Also referred to as robber crabs, Birgus latro is closely related to the hermit crab, and in fact goes through a hermit phase. Adults are about a meter long from head to tail and weigh about 17 kg. Each has a set of two large pincers that constitute its primary tool for acquiring and preparing food. The food most often commented on is the coconut, Birgus is reputed to be able to tear the husks off nuts, pierce their hard inner shells, and then crack them open to extract the meat, but we can find no published support for this reputation. There is abundant evidence, however, including our own observations, that coconut crabs are omnivores, and readily consume meat. We observed them eating rats, delicately peeling the flesh out of the skin, which was discarded. Whether the rats were carrion or were caught alive is something we cannot say, and on which the literature appears to be silent.

Coconut crabs are solitary, though they may be so numerous in some places that they appear to be gregarious. They tend to move along "paths of least resistance" between the shore and interior areas, particularly during the breeding season (roughly December to February). After copulating on land, females make their way to the shore, where they release their larvae into the water. The larvae go through a planktonic stage and then come ashore as very small crablets. In this phase they begin carrying and hiding in the abandoned shells of univalves. This juvenile shell carrying phase continues until a crab is about two years old, with a carapace about two to three centimeters across. At this point the crab abandons its shell carrying practice. By the age of five years the crab is about ten centimeters across, and they continue to grow, apparently, until death. Coconut crabs have been known to live for over forty years.

Coconut crabs dig burrows in which to moult. Burrows may be complicated tunnel systems ending in chambers, usually not over 50 cm. below the surface. Crabs have been observed to remain in their burrows for up to sixteen weeks, shedding the old exoskeleton and hardening the new one.

Coconut crabs abound on Nikumaroro. Niue Islanders who worked on the island in the late 19th century referred to it as Motu Aonga--the land of the coconut crabs, and in late 1937 a party of I Kiribati and British explorers built fires in a ring around its campsite--"as protection against the giant robber crabs, who stalked about in the half-light or hung to the branches staring balefully at us."

Beachrock Ramparts

Nikmaroro is a apparently a bit different from other atolls in the area on account of its beachrock. These ramparts, apparently formed in the early Holocene (10ky), have given the island a terrestrial rockmass core (above current sea level) that has trapped sediment, either aeolian or overwash. These sediments in turn become more beachrock as Ca-saturated groundwater from the lagoon passes through them, cementing the particles together.

When an atoll is able to do this, it seems to acquire a stable core, while others that do not have beachrock, don't (e.g. Carondelet Reef). Beachrock cores to atolls would be a fascinating study as they offer a protection to the atoll that is otherwise absent ("rampart" is the name I have given the eroded beachrock outcrops that are above current MHWS). They would be what could protect an atoll from sea level rise - colonise them that do have ramparts, abandon them that don't.

Howard Alldred NZCS BSc(Hons)

Geodata Consulting Ltd. Waiheke Island New Zealand

Oceandots Synopsis

Measuring around 6 km in length and up to 2 km across and with a land area of 4.1 km² the atoll forms an elongated structure enclosing a narrow lagoon. There are two passes that split the rim into three sections — the narrow Bauareke Passage on the south side and the broader Tatiman Passage at the western end. Note that neither of the passes cuts through the broad reef flat that completely surrounds the atoll. Much of the land is thickly vegetated with Pisonia grandis, Cordia subcordata and Cocus nucifera.
Like all of the Rawaki Islands, the north, south and eastern sides of the atoll are pounded by heavy surf and are exposed to storms and Pacific swells from the east and south (note the thickness of the surf-line around three sides of the island in the image).