lacked Pandanus when we surveyed the island. Moreover, the 

 first islet south of Nake supports an extensive grove of large 

 Pandanus trees on its eastern (lagoon) shore, and we feel 

 confident in ascribing Long's specimen to this island, which 

 we had named "Pandanus" because of this grove. To be 

 consistent, we have ascribed all his other "second islet" 

 specimens to Pandanus Islet as well and assume he made no 

 collections on the actual second islet (Danger). 



Species Lists. Annotated Checklist, and Maps of Terrestrial 

 Vascular Plants : Following recent authors Sachet & Fosberg, 

 19S3;Lamberson, 1987), we do not considerCaroline's transient 

 or extinct vascular flora (Table 1) or the vegetables and 

 ornamentals in the Falconer's garden as part of Caroline's 

 viable flora. Table 2 summarizes the current flora, detailing the 

 relative abundance of each species within each plant community . 

 These tables are based on sight records supplemented by all 

 collections, past and present. No beach drift seeds are known 

 from Caroline apart from those species already represented. 

 English and Gilbertese names in Tables 1 and 2 are from 

 Thaman ( 1987), St. John (1973). and Perry & Garnett (n.d.). 

 If no common name is available, the Hawaiian name, familiar 

 to many students of Pacific botany, is used. 



One new species record for the atoll, as yet unidentified, is 

 called Species A (K-90-23. 24): a single sterile shrub 2.5 in 

 high, collected from southwest Motu Mannikiba in coarse 

 strand rubble, is similarto Clerodendrum inerme, with leathery 

 leaves and arching stems 4-5 m long. It was found by John 

 Phillips. 



Table 3 lists the distribution and abundance of plant 

 species ( with subdivisions into tree, shrub, and herb components) 

 onallmotus. Figures 1 1-25 map the entire atoll distribution of 

 each species according to data from transects and aerial maps. 

 Families are arranged phylogenetically. according to 

 losherg & Sachet ( 1 987 ). with species arranged alphabetically 

 within each family. The taxonomy of vascular plants follows 

 W. Wagner etal. ( 1990). and ferns follow H. Wagner (personal 

 communication). "'< cover" means the percentage of the 

 ground area covered by a particular plant species. In all text and 

 tables, the following symbols apply: 

 New record for Caroline 



Indigenous — plants native to Caroline but also occurring 

 elsewhere (I) 



Aboriginal introduction — useful plants brought by 



Polynesians in pre-historical times (AI) 



# Recent introduction — plants of accidental or deliberate 



introduction alter Western discovery of the atoll (Rl) 



A Abundant — generally the major or dominant species in 



a given area 

 VC Very common — often seen but not quite as abundantly 



as above 

 C Common — generally distributed throughout a given 



area in large numbers 

 UC Uncommon — observed uncommonly but more than 10 



times in a given area 

 Occasional — here and there, often widely scattered but 



not forming a major component of the vegetation 

 R Rare — observed 2 10 times in a given area 

 S Single — only one specimen observed 



L Local — found only or principally in one or more restricted 



areas 

 D Drift seedling — plant derived from a water-borne seed 

 + Not seen 1988-1990 but probably still present 



PSILOTACEAE 



* Psilotum nudum (L.) Beauv (Fig. 1 1 ) 



Formerly Known Distribution : L-3233 from Nake 

 Present Distribution : Cosmopolitan, common on remote 

 islands, rare on Caroline. K-90-15 from South. In 1965. 

 common on wet base of Cocos only on Nake Island. In 1988 

 and 1990. a few clumps found similarly on South Island in 

 shady, damp locations, close to lagoon, northwest sector. 

 Cocos canopy was 18 m. 



POLYPODIACEAE 



* Phymatosorus scolopendria (Burm. /) Pichi-Sermolli 

 (Fig. 12: PI. 34) 



Phymatodes scolopendria (Burm./) Ching 



Polypodium phymatodes L. 



Polypodium scolopendrium Burm./. 



Microsorium scolopendria (Burm.) Pichi-Sermolli 

 Formerly Known Distribution : Recorded 1840. collected 

 1884; L-3244, L-3250, L-3287 from Nake, Long, and South 

 Islands. 



Present Distribution : Range extension from 3 to 1 1 motus. 

 Rarely a continuous ground cover, usually locally rare to 

 abundant. Commonest on Nake. with cover 10-80%. Well 

 represented on South, especially in open areas of the interior, 

 where soils are moister. On other motus local distribution 

 varied from less than 1 to 80%; accurate mapping is difficult. 

 Absent from motus less than 0.6 ha in size, where habitats 

 cannot provide cover, moisture, and substrate for both 

 sporophyte and gametophyte generations. 



Ecology : Hardy. Leaves burn in sun but can withstand 

 very dry conditions. Primarily in Toumefortia scrub, mixed 

 forests with Pisonia and Pandanus. or Cocos plantations. 

 Associated with Cordia, Morinda, Suriana. In open clearings 

 within dying Cocos forests, occurs in dense mats intermingled 

 with Boerhavia, Ipomoea, and Portulaca. Sometimes gathers 

 in thick bands at the interface of Toumefortia and Pisonia 

 forests. Prefers shelter, high humidity, "soil." and relative lack 

 of wind, but absent from deeply shaded forests. Rhizomes 

 never exposed on ground surface or epiphytic on trunks, as in 

 wetter islands such as Hawaii or Samoa (personal observation) 

 or m the moister Line Islands (Wester. 1985). indicating that 

 Caroline's habitats are suboptimal. Although most ferns are 

 not halophytic, tins species grew (rather stunted) in l l >NN 

 amongst sparse herb mats ( 1% cover) on older beach sands of 

 an ancient reef channel on Long Island (Tr. C). where rainfall 

 pi ov ides the sole fresh water, but was ( temporarily?) obliterated 

 in February 1990. Rare to uncommon in outer beach strand, 

 and beach scrub w ith Suriana on South, Arundel, and Shark. 

 Substrata: Dry coral rubble, sand and gravel, rubble with 

 sparse humus, lagoon mud, relatively fertile humus, older 

 beach sands. 



