The Soils of Kiritimati (Christmas) Island, Kiribati, Central Pacific: New Information and Comparison with Previous Studies1 R. J. Morrison2,3 and C. D. Woodroffe3,4 Abstract: Kiritimati, the largest land area atoll in the world, is undergoing rapid population increase, and, given the isolation of the island, local food production will have to be expanded to support the residents. Two soils investigations were completed in the 1960s, but no additional information on the soil resources of the island has been produced since that time. In this study, 15 soil profiles were described and analyzed. Where possible, comparison has been made with previous work, and discussion of the soil-forming factors is presented. Results confirm that soils are weakly developed (Entisols) with relatively low organic matter contents and low water-retention capacity. These properties are expected from the age of the parent materials and the relatively dry climate of the island. Total elemental analyses show that the soils contain very low concentrations of potassium and important trace elements (iron, manganese, copper, and zinc), which will limit any plant production. Classification of the soils identified eight soil families, mainly separated on the basis of content of larger coarse fragments and soil moisture regime, including the influence of groundwater. Comparison with previous studies showed that although different nomenclature and classifi- cation systems were used, similar soil patterns were observed, and the soils of Kiritimati are relatively unique in the Pacific islands. Kiritimati, located at 2 00 0 N, 157 30 0 one of the most highly populated single-atoll W, is the largest atoll in the world in terms of islands. The population has risen markedly land area (@320 km2). The island was ‘‘dis- since the early 1990s, when the government covered’’ by Captain Cook on 24 December of Kiribati started encouraging people to 1777 and was named Christmas Island. It is move to Kiribati to reduce population pres- located at the southern end of the Line Is- sure on the Gilbert group, especially Tarawa. lands, which have a pronounced north-south Such an increase will have impacts on the nat- wet-dry rainfall regime, with Kiritimati lying ural resources of Kiritimati, especially its soils in the equatorial dry zone (Schott 1933), a and water. narrow belt of low rainfall across the eastern The island is shaped somewhat like a lob- central Pacific Ocean. Kiritimati has a popu- ster claw (Figure 1) with a lagoon covering lation (2005 census) of over 5,000 (Secretariat some 160 km2 occupying much of the west- of the Pacific Commission 2005), making it ern end of the island. East of the main lagoon is a central area containing more than 500 saline or hypersaline lagoons covering some 1 This work was supported by the Australia Pacific 140 km2 that have been the subject of several Science Foundation. Manuscript accepted 16 August studies (Valencia 1977, Trichet et al. 2001, 2008. Saenger et al. 2006) because they present rel- 2 Corresponding author (e-mail: [email protected]). 3 GeoQuEST Research Centre, School of Earth atively unique environments for algal growth. and Environmental Sciences, University of Wollongong, There is a limited tourism industry based on NSW2522, Australia. recreational fishing and bird watching, but 4 (e-mail: [email protected]). copra production is still a major component of the local economy. Local food production Pacific Science (2009), vol. 63, no. 3:397–411 is now also a priority, and knowledge of the : 2009 by University of Hawai‘i Press soils is essential for sustaining the local popu- All rights reserved lation. 397 398 PACIFIC SCIENCE . July 2009 Figure 1. Map of Kiritimati showing locations of the soil profiles studied. Information on the soils of Kiritimati is Previous Studies on Soils and Environment of substantially better than that for most atolls Kiritimati in the Pacific as a result of two studies completed in the 1960s by Jenkin and Foale Early comments on the soils of Kiritimati (1968) and Hammond (1969), which are com- can be found in Christopherson (1927) and mented on further in the next section. Those Wentworth (1931), but they were mainly in- reports are now over 30 yr old, and this study terested in the vegetation and geology of the was initiated to make available some more- island. Although there was a spurt of interest current information for decision makers on in atoll soils in the post-World War II era the future of Kiritimati. This study involved (e.g., Stone 1951, Fosberg 1954), limited at- a site visit to the island; description, sampling, tention was paid to the ‘‘dry’’ atolls until the and analysis of key soils; and comparison with 1960s, when two notable studies on Kiritimati the results of previous investigations. were completed by Jenkin and Foale (1968) Kiritimati New Soils Information . Morrison and Woodroffe 399 and Hammond (1969). The former was an in- the soils were considered phases of the vestigation of the coconut-growing potential Shioya soil series. This was based on similar- of the island, and the latter was a more tradi- ity in profiles with the Shioya soils originally tional pedological study. These two studies mapped on Okinawa but found also in Arno produced valuable information but were (Stone 1951) and the northern Marshall Is- somewhat limited by the analytical facilities lands (Fosberg and Carroll 1965). As noted available at the time. by Morrison (1990), there is an issue with in- Jenkin and Foale (1968) divided the soils cluding these Kiritimati soils in the same se- into two associations depending on the depo- ries as those from the wetter western Pacific sitional mode of the patent materials: oceanic islands; they would be separated at the great deposits and lagoon deposits. The oceanic group level in Soil Taxonomy (Soil Survey deposits were further differentiated on the Staff 1999) based on the prevailing moisture basis of particle size (sands, mixed coral regime because the Shioya series soils were sand þ gravel þ rubble, and coral rubble), previously mapped on higher-rainfall loca- with the coral sands being separated into tions. Six mapping units were included in three types based on location and associated Hammond’s (1969) soil map (the Shioya vegetation type: young beach sands, coastal gravelly phase was of insufficient distribution dune sands, and older beach sands. The to map separately). Organic matter contents lagoon-deposit soils were differentiated on were low, C:N ratios average about 12, and the basis of the presence or absence of a the pHs were high (>7.7). Potassium con- hardpan. Laboratory information obtained centrations were very low, and most of the included pH (in H2O and CaCl2 ), % CaCO3, phosphorus was found in the surface hori- % organic C, % N, total K, P2O5, Mg, Na, zons; no trace element information was col- and Sr. Attempts were made to measure trace lected. elements (e.g., Cu, Fe, Mn, Mo, V, and Zn), but in almost all cases the results were below Soil-Forming Factors on Kiritimati the detection limits available at the time. Low concentrations of trace elements are common climate. As noted earlier, Kiritimati is in in atoll soils derived from coral limestone the equatorial dry zone and receives substan- having a relatively high pH (Morrison 1992). tially lower rainfall than islands to the north, Jenkin and Foale (1968) did not complete any south, and west. The average annual rainfall detailed classification of the soils but consid- is 914 mm (based on data from 1951 to ered them as Entisols, Lithosols, and Rego- 2006) but is extremely variable, ranging from sols, because they showed minimal profile 177 mm in 1954 to 2,959 mm in 1987 (Figure development. 2). These large variations are linked to varia- Hammond (1969) examined and described tions in the El Nin˜ o–Southern Oscillation about 30 pedons around the island. He also (ENSO) phenomenon, with high rainfalls completed laboratory analyses including pet- occurring during El Nin˜ o episodes and rographic examination, mineralogy by X-ray droughts in the intervening periods (Falkland diffraction, particle size distribution, major and Woodroffe 1997). Falkland and Wood- total element analysis (Ca, Mg, P2O5, Na, K, roffe (1997) reported that there is a strong and SiO2), % organic C, % N, pH, and elec- correlation between annual rainfall and the trical conductivity. He found minimal change Southern Oscillation Index (SOI). Evapo- in the total element chemistry between the transpiration is high ( Jenkin and Foale soils and the parent materials, confirming 1968), with estimates often exceeding annual their youthfulness. He divided the soils into rainfall, confirming the relative aridity of the seven categories, separated on the basis of island. The applicability of some of the equa- landform position, particle size distribution, tions to calculate potential evapotranspiration and hardpan presence/absence. Except for in a situation like Kiritimati is questionable one category, unaltered sand and gravel, all given the high porosity of many of the soils, 400 PACIFIC SCIENCE . July 2009 Figure 2. Annual rainfall (mm/yr) for Kiritimati, 1951–2006. with water moving through the profiles be- where Foraminifera are the major sediment fore evaporation could occur. source. The mean annual temperature on Kiriti- In terms of particle size the parent materi- mati is 27.5C, with limited variation (mean als are dominated by sands, with the quanti- monthly maximum about 30.3C and mean ties for coarser material (>2 mm) varying minimum about 24.6C [Kiribati Meteoro- with location. The proportions of very fine logical Service, pers. comm.]). Winds are sand þ silt þ clay are always less than 5%. predominantly easterlies, but infrequent relief/landform/geology. The mor- northeasterlies are important because they phology of Kiritimati reflects that of the are the main winds bringing rain.