1 The objective of this study was to explore the interrelationship bet
ween island area, number of species, and habitat diversity. 2 A survey
of dicotyledonous plant species was carried out on 45 uninhabited, un
improved, small islands off Shetland Mainland, plus two similar mainla
nd headlands treated as islands. In addition, species were counted wit
hin 50-cm square quadrats randomly placed on island vegetation. The la
rgest island surveyed was 100 ha; 81 plant species in all were found.
3 A total of 14 physical (abiotic) habitat types were classified. The
number of habitats on each island was counted, and the habitat types c
haracteristic of each plant species were recorded. Island areas were d
etermined from Ordnance Survey maps. 4 There are close-fitting positiv
e correlations between species number, island area, and the number of
habitat types on an island. 5 Data on species number within quadrats o
f standard area reveal an increase in small-scale species richness on
islands of increasing size - evidence for an effect of island area alo
ne on species total. 6 Habitat types containing fresh water were large
ly absent from islands of less than one hectare in size. Species prima
rily associated with fresh-water habitats were generally also missing
from these smaller islands - evidence for an effect of addition of hab
itat types on species total. 7 Path analysis confirms that island area
contributes to species number both directly and indirectly, through h
abitat diversity, and that while the direct effects of area and habita
ts on species are roughly equal in magnitude, the total effect of area
is nearly twice that of habitats. 8 Presence and absence of particula
r habitat types may be a function of island size.