This work investigates 2 US Army Construction Engineering Research Laborato
ry revegetation experiment sites (Phase I and II) on Kaho'olawe Island, Haw
ai'i (USA) to determine the longterm success of several revegetation strate
gies and to identify species that are best suited for future restoration ac
tivities in the highly eroded central plateau region of the island. Only th
e Phase I treatments receiving the highest rates of phosphorus and nitrogen
fertilizer and the Phase II strategy (moderately high fertilizion, and lan
dscaping) produced enough cover to begin providing protection from erosion
processes. Buffelgrass (Cenchrus ciliaris L,), glycine (Neonotonia wightii
(Wight & Arnott) Verdc.), Natal redtop (Rhynchelytrum repens (Willd,) Hubb,
), and siratro (Macroptilium atropurpureum (DC) Urb,) appear to be resilien
t to the harsh island conditions, which include strong winds, low annual ra
infall, acute erosion, and a nutrient-depleted soil profile. Although all 4
species offer some protection against erosional processes, none are partic
ularly desirable for long-term restoration of the island. Glycine and sirat
ro, both of which volunteer readily in planting sites, are considered invas
ive in that they may smother other more desirable species, particularly les
s competitive natives. Finally, native woody species are shown to have diff
iculty in surviving on the island without special attention to planting and
maintenance.