Pandanus pistillaris in the Caroline Islands: 
An Example of Long-Range Oceanic Dispersal 
Benjamin C. Stone 1 
In CONJUNCTION with the current revision of 
the genus Pandanus (St. John, I960; St. John 
and Stone, in sched.) it seems appropriate to 
mention the factors of dispersal operative in 
the genus. Since a number of species of Pan- 
danus are littoral in habitat, and because at least 
one (if not several) species is found on nearly 
every tropical atoll in the Pacific, it has rightly 
been assumed that oceanic dispersal of the buoy- 
ant fruits is responsible for the distribution. 
There is no question that throughout vast island 
groups in Micronesia, Melanesia, and Polynesia 
Pandanus has long been part of the indigenous 
vegetation. There are, in fact, some indications 
that oceanic dispersal accounts for the farthest 
limits of the genus, in West Africa at one ex- 
treme, Polynesia at the other, the Bonin Islands 
in the north, and Australia in the south. The 
actual number of species involved is not yet 
known, but species of section Pandanus ( section 
Keura of earlier authors; see St. John, I960) 
play a prominent role, especially in the Pacific 
region. Certain species of other sections of the 
genus also are primarily ocean-distributed (for 
example, Pandanus duhius Spreng., and Panda- 
nus tetrodon ( Gaud. ) Balf i. ) . In general, fruits 
of these species which are buoyant in seawater 
exhibit anatomical structures which are pre- 
sumably adaptations to permit, or prolong, flo- 
tation. Brown (1931) has discussed the tissues 
present in fruits of certain Polynesian species, 
and uses the apt term aerenchyma to designate 
the light, pithy, upper mesocarp of such fruits. 
In addition, the seeds are encased in a highly 
1 Department of Botany, U. S. National Museum, 
Smithsonian Institution, Washington 25, D. C. This 
work was carried out while the author was Research 
Assistant (under N. S. F. Grant 1834) at the Univer- 
sity of Hawaii, Dr. H. St. John, principal investigator. 
Grateful acknowledgment is expressed to Dr. St. John 
for his aid, and to Dr. M. S. Doty and Dr. T. F. Aus- 
tin, of the University of Hawaii, and Y. Sinoto of the 
B. P. Bishop Museum, for their criticism of this con- 
tribution. Manuscript received January 9, 1961. 
sclerified osseous endocarp which, except for the 
minute apertures at either end, are nearly imper- 
vious to water. Fruits of this type are capable of 
remaining afloat for a considerable period. No 
experiments or observations seem to be availa- 
ble, but it would appear quite probable that a 
period of 6 months or 1 year in the ocean would 
not be an exaggerated estimate of the time 
in which floating fruits could be dispersed by 
currents. 
The vast areas encompassed by some of these 
species or groups of species with buoyant fruits, 
and the considerable taxonomic difficulties en- 
countered in the groups, have not heretofore 
permitted any definite observations to be made 
with respect to direction of dispersal. Over a 
period of many years, and because of numerous 
observations, the knowledge of the currents in 
the Pacific is now in a position to be of value 
in problems of oceanic dispersal; but when grave 
doubt exists as to the nature (and number) of 
entities involved, little can be said except the 
obvious, that Pandanus species of littoral areas 
are probably ocean dispersed. 
The specimen discussed below, however, be- 
cause of its unique character, is a reasonably ac- 
curate indication of one route of oceanic disper- 
sal. Because it represents a species which is a 
member of section Intraob tutus St. John, it is at 
once distinguishable from the widely dispersed 
members of section Pandanus . 
The specimen was collected by Dr. Ryozo 
Kanehira, late professor at Kyushu University, 
Fukuoka, Japan; the label reads, 'Mokil Atoll; 
March 12, 1937, Kanehira no. 4203; nom. ver- 
nac. "arowan.” Mokil Atoll is an isolated atoll 
roughly 110 mi. east of Ponape, at 6° 40' N., 
159° 45' E., consisting of three islands on a cir- 
cular reef. In April 1957 the present writer 
visited Mokil, spending the day of the 16th bo- 
tanizing on the major islet. During this period, 
15 numbers of Pandanus were collected (all of 
section Pandanus) and many vernacular names 
610 
