In September 1988 we saw no flowers or fruit. Anne 

 Falconer reported flowers on Motu Ana- Ana in August 1990. 

 Pisonia was beginning to bloom on Vostok in March 1990 

 (A. Kepler, in prep.). 



A Historical Perspective : Some of Caroline's most mature 

 Pisonia groves (to 21 m tall, 660 cm circumference at 1.5 m. 

 multiple trunks) appear to be virgin. Overall dimensions of the 

 trees, low species diversity, and general character of the plant 

 community are similar to the known virgin groves on Vostok 

 (personal observation, Table 12). We do not have dimensional 

 data (other than height) of these particular areas on Nake, but 

 their level of maturity can be seen in PI. 43. 



Despite the advanced stage of ecological succession of 

 many groves, especially to windward, planting records from 

 1916-1922 indicate that Cocos was planted throughout, not 

 only South, but also on Nake. Long, and all the major Windward 

 Islets (Young, ca. 1922). Given the standard planting density 

 of one tree every 8.5 nr (28 x 28 ft) ( Young, ca. 1922), we have 

 calculated the approximate area on each islet given over to 

 Cocos plantations, based on the number of coconuts planted 

 (Table 13) times the area required for each tree (73 nr). We 

 have then compared this to the usable areas based on today's 

 forest cover. On all nine Windward Islets, Cocos covered 

 79 to 100% of usable ground; in several cases the amount 

 calculated for Cocos exceeds the amount of potentially usable 

 ground. Thus, Cocos was so intensively cultivated on the 

 Windward Islets that essentially all the Pisonia and most 

 Toumefortia forests must have been felled. 



Two remarkable points emerge from Table 13: 1. Scarcely 

 any Cocos remains today on the nine Windward Islets; seven 

 of them bear no trace of the former plantations 

 (Figs. 43,44,47,48); and 2. recovery of Caroline's natural 

 ecosystems. Stages I through V ( Ecological Succession section ). 

 on the windward side has been rapid and, at least on Brothers 

 Islet (Fig. 46), reasonably complete with regard to ecological 

 succession and species diversity. Today the Windward Islets 

 have the lushest and tallest plant communities, with higher 

 species diversity than the leeward islets (Table 3). which have 

 evidently experienced far less human disturbance. 



This differential disturbance on the w indward and leeward 

 sides of the atoll explains enigmas such as 20-m-tall Pisonia 

 forest on the leeward Booby Islet (0.84 ha), taller than most of 

 the windward forests; the absence of Pisonia on windward 

 Tridacna Islet (9.08 ha), which, being close to South Island, 

 probably supported Cocos, which was managed longer than the 

 more distant windward islets; and the patchy distribution of 

 Pisonia in the interiorof several islets (e.g.. Windward, Arundel). 

 This last point also applies to Mannikiba ( 2 1 .49 ha), the largest 

 leeward islet. According to Young (ca. 1922), 6,000 seed sets 

 were brought from Flint to Caroline in 1920 and kept on 

 Mannikiba. This "nursery stock" was used to replant "misses" 

 on other islets, due mostly to destruction by coconut crabs and 

 poor planting. Today. Mannikiba's total acreage of Pisonia 

 (Fig. 53) is very small and fragmented relative to the islet's 

 si/e: 1.13 ha. 5'< of the total land area. Compare this with Bird 

 Islet (Fig. 55). which, as far as we know, has never been 

 disturbed: 1.70 ha Pisonia, 42' i of the islet's land area. 



On both Caroline and Flint there is much variation in the 

 quality of the regenerated Pisonia forests (Table 12). Some 

 trees bear enormous, partly rotting boles, black algae smothering 

 the bark, multiple trunks, and few or no understory herbs. 

 Other trees are tall and straight-trunked, with characteristic 

 whitish bark, and bear no rotting holes in their bases. Such 

 observations suggest that when their indigenous forests were 

 felled, only minimal cutting was done, and many Pisonias were 

 able to regenerate quickly by sprouting from rooted stumps and 

 fallen branches. This speculation is supported by the fact that 

 some of Vostok's Pisonia trees regenerated similarly. Maude 

 (1953, p. 96) stated that "there is room for 8,000 palms on 

 Vostok, but only 100 have been planted and most of these have 

 been choked in the luxuriant 'buka' (Pisonia grandis) forest: 

 no attempt having been made to exploit the island since the 

 initial planting." 



Pisonia, a soft, pulpy wood, has a well-known ability to 

 sprout or send up suckers from dismembered branches or fallen 

 trunks ( Fosberg, 1 953), and it has been noted that older trees are 

 virtually indestructible, fire being the only effective means of 

 clearing forests (Wiens, 1962, p. 397). The senior author has 

 photographed leaf sprouts from partly burned twigs as small as 

 1 m long and 5 to 6 cm in diameter. 



Since the existing Cocos plantations on South Island and 

 southwest Nake contain few Pisonias, it seems that forest 

 clearing was more thorough on the atoll's larger islets than on 

 the smaller ones, which today manifest scant traces of their 

 former history. Fortunately for Caroline, its coconut plantations 

 were plagued by a number of problems, which resulted in their 

 being abandoned twice: coconut crabs, seabirds, rats, 

 Ipomoea vines, and an unknown disease (see under Coconut 

 Woodlands, this section). 



A footnote in Young (ca. 1922, p. 15) stated that "the larger 

 portion of the 30,000 trees planted were either badly planted or 

 smitten with some disease as in 1927 it was reported by 

 Mr. Bunckley that most of them had perished." In 1929 only 

 1 3,2 1 5 trees were left and more were being planted. Considering 

 the distribution of both palms and natural forests today, it 

 appears that plantations continued on South and Nake and were 

 abandoned on the smaller islets, allowing for a better recovery 

 than might be expected had the Cocos grown to maturity. 



Once a Cocos plantation has been well established and 

 subsequently abandoned, Pisonia regrowth is more difficult. 

 This is characteristic of mans islands in the tropics. For 

 example, on Cousin Island (an ICBP wildlife preserve since 

 1968, Seychelles Islands. Indian Ocean), where Pisonia is 

 currently reestablishing within adeterioratingCf)cc.v plantation, 

 Phillips & Phillips ( 1990. p. 37) envisioned "centuries rather 

 than decades before something like a natural ecosystem 

 develops." We predict a similar time frame for areas on 

 Caroline and Flint where Cocos canopy is over 7095 cover. 



Annual Growth Rates : Data on Pisonia grandis growth 

 rales are evidently lacking. On Cousin Island, vegetation 

 changes, including Pisonia and Cocos, have been monitored 

 since 1 474 (Phillips. 1984; Phillips & Phillips. 1990). However. 

 as in) measurements of any tree dimensions are included, 

 growth rates cannot be deduced. 



30 



