discovery" ( 1606), de Quiros and his party were desperate for 

 fresh water. After noting how lush and green Caroline was, 

 they expected to find good water supplies, but there was 

 "nothing but salt water in the holes they dug" (Markham, 

 1904). Maude ( 1968) suggested, in retrospect, that had they 

 waited longer the salt water in their shallow wells might have 

 run fresh, as has been his experience on some other atolls. 

 During the 19th century, three wells were used — one on Nake 

 and two on South (Holden & Qualtrough, 1884). One South 

 Island well contained fresh water at 1.5 m depth in 1974 

 (Garnett. 1983). We saw no wells but located concrete cisterns, 

 one built near the northwest point of South in 1937 and rebuilt 

 by the Falconers in 1989. and another uncovered one (dating 

 from 1938) within a Cocos—Pisonia grove along Tr. 2, about 

 200 in east of the southwest corner of Nake. 



Caroline's paucity of fresh water may be partly responsible 

 for the lack of a permanent population. The annual rainfall in 

 1989 (Appendix 2) was 1,242 mm (48.9"). However, like the 

 similarly lush Nikumaroro and Orona (Phoenix Islands), 

 Caroline's rainfall may vary greatly from yearto year, resulting 

 in undependable water supplies. In the past, residents relied on 

 rainfall catchment for fresh water (Maude, ca. 1938; R. Falconer, 

 personal communication). 



Shallow sources of fresh or brackish water may be present 

 on most islets, as Pisonia forests occur on 14% of them, but 

 very little is known of freshwater lenses supporting Pisonia 

 forests (Wiens, 1962). If we assume that Pisonia is not salt- 

 tolerant, small water lenses may be present on motus as small 

 as 0.2 ha. This is further discussed in the Ecological Succession 

 section. 



Climate 



Meteorological records for Caroline were sparse until 

 1989, when Ron Falconer began daily records of rainfall and 

 wind direction (Appendix 2). Some data is available from the 

 plantation years 1916-1920 (Young, ca. 1922) and during the 

 1883 Solar Eclipse Expedition (20 April-8 May) (Upton. 

 1884), when 203.2 mm (8") fell. The best generalizations on 

 weather conditions in this area are found in the Geographical 

 Handbook Series: Pacific Islands, Volume 2 (N.I.D.. 1943). 

 Wiens (1962), Seelye (1950), Taylor (1973), and various 

 papers on the Tuamotus (Stoddart & Sachet, 1969; Sachet, 

 1983). Islands in the Line Group experience a wide range of 

 climates. In general, those near the equator are dry, with 

 rainfall increasing with increasing latitude north or south. 



Caroline experiences a tropical oceanic climate that varies 

 little during the year. Temperatures are uniformly warm to hot, 

 normally tempered by trade winds from the southeast to 

 northeast. Falconer (personal communication) has recorded an 

 annual average of 30°C (86°F) (range 26-3PC [78-88°F|). 

 Mean annual temperatures for the Central Equatorial Islands 

 lie between 24 and 29 C (75-85°F). Surface temperatures 

 increase rapidly in early morning and remain hot throughout 

 the day. forest interiors are humid. The daily range of 

 temperatures exceeds the annual fluctuation in the daily mean. 



Atmospheric pressure, sunshine, and cloud cover are 

 probably similar lo the northern Tuamotus — uniform except 



during storms. During our 1 988 surveys the sky and air were 

 extremely clear, but we experienced much heavy cloud and 

 rain in March 1 990, the aftermath of cyclones "Peni" (centered 

 near Vostok) and "Ofu" (centered further west). 



Wind and Rainfall : Caroline is dominated by trade winds. 

 As on all low atolls, land topography has no appreciable effect 

 on weather. Although it lies within an area primarily influenced 

 by southeast trades, there is a small annual oscillatory movement 

 northward and southward, so that in reality winds blow from 

 the east, northeast, and southeast. This accounts for the east 

 and northeast winds that puzzled the eclipse party, who were 

 expecting winds from the southeast (Upton. 1884). Data from 

 Falconer (Appendix 2) indicates that, at least for 1989 and 

 1990, winds blow primarily from the north and northeast, and 

 rarely from the southeast (April-August). 



The atoll lies within a belt of variable rainfall, along with 

 Vostok, Flint, and the northern Tuamotus. Young (ca. 1922, 

 p. 13) notes that Caroline's rainfall is "certainly less than that 

 of Flint," giving exact figures for 1919 (2,172 mm) and 1920 

 ( 1.854 mm) and estimates that there was "probably not more 

 than 50" (1270 mm)" during 1916. 1917. and 1919. These 

 estimates were based on exact figures from Flint ( 1 .600. 1 .346, 

 1,295 mm, respectively). Falconer measured 1.242.1 mm 

 (48.9") in 1989 and 2,209.8 mm (87") in 1990. An unusually 

 stormy February in 1990 brought 640 mm (25.2") of rain. 

 Rainfall distribution isohyets (Taylor. 1973) assign Caroline 

 an annual precipitation of approximately 1,500 mm (60"), a 

 perfect average ofthe above 6 years (x= 1,513mm). Ingeneral, 

 "winter" (May-October) corresponds roughly to a dry season 

 and "summer" (November-April) to a wet season. 



Hurricanes and Tsunamis : Atoll islets are active structures, 

 undergoing repeated death and rebirth. Violent storms contribute 

 to ongoing erosional and rebuilding processes. Storms deposit 

 debris not only along the shores of the windward islets 

 (Pis. 17,20) and across reef flats into the lagoon but sweeps it 

 far inland. 



Although the south-central Pacific is relatively free of 

 cyclonic storms (cyclones, typhoons, hurricanes), they do 

 occur with enough frequency and devastating force that any 

 discussion on climate should include them. Although detailed 

 records of hurricanes and tropical storms exist for the inhabited 

 Tuamotus since European discovery, many of these may not 

 have affected Caroline. However, the following evidence 

 suggests that Caroline experienced two major hurricanes last 

 century and that periodic violent storms can modify the atoll 

 substantially: 



/. Between IH22 and 1825. When de Quiros visited 

 Caroline in 1 606. the northwesterly Cocos plantation on South 

 Island was healthy. When Bennett arrived in 1834. he noted 

 that all the palms were "of dwarf stature." and that "amidst the 

 original groves, the number of vigorous seedlings fully 

 confirmed Captain Stavers' statement [who had visited the 

 atoll in 1828] that these palms had increased greatlv since his 

 last visit to the spot" (Bennett. 1840). 



A few years before 1 828. therefore, something had affected 

 the palms. By 1834 they were all of an even height and quite 

 short, vet bore nuts. French records indicate that two devastating 



14 



