88 
PACIFIC SCIENCE, Vol. XV, January 1961 
All other land animals collected or observed by 
me, or those later sent me from the weather 
station, are listed in Table 3, accompanied by 
notes on distribution and abundance. The fol- 
lowing remarks will explain the data presented 
in the table. 
1. The number of species given in parentheses 
after the names of the major taxonomic groups 
does not always coincide with the number listed 
under each group, because the familial char- 
acters were not determined for some specimens 
which were not caught or were lost before or 
during the preparation of our data. 
2. Under the column showing range, the dis- 
tribution of identified species in other districts 
is mentioned. In the majority, however, only 
the pattern of geographic distribution is given, 
using the following abbreviations: E, endemic 
at present; C, cosmopolitan; T, tropicopolitan, 
including Indo-Pacificopolitan; P, Pacificopoli- 
tan; and Pa, Palaearctic. These patterns are 
naturally very conventional, for the distinction 
among C, T, and P is often subjective. 
3. Under the column showing abundance, the 
relative abundance of each species is shown with 
marks: + + , very abundant; ±, abundant; +, 
common; — , rare. The last observation may 
express not an actual rareness, but only a cryp- 
tic life-mode. 
4. The distribution and relative abundance of 
each species in the various habitats (see the 
definition of A, B, etc., in the next paragraph) 
are indicated by O (for occurrence) or A (for 
abundance). Where holometabolic insects are 
concerned, the distribution is considered only 
with respect to adults, but in the sphingid and 
noctuid moths, only with respect to their cater- 
pillars, inasmuch as the adults were collected 
only at lights. 
5. The species observed but not collected are 
marked with an asterisk, and those which were 
only indirectly confirmed are marked with a 
dagger. 
DISTRIBUTION OF ANIMALS 
IN VARIOUS HABITATS 
In order to obtain a closer perspective with 
regard to the ecological distribution of animals 
listed above, the island was divided into the 
following habitat zones, based upon topography 
and vegetation (Fig. 12). 
A: Areas with both arboreal and herbaceous 
strata (Fig. 4 and Fig. 5, back) : 
Ai : Floor stratum, including earth surface 
and sites beneath gravel and stones. 
A 2 : Herbaceous stratum, consisting of Ipo- 
moea foliage alone. 
A 3 : Arboreal stratum, consisting of Mes- 
serschmidia and Pisonia as dominants. 
B: Areas without arboreal stratum, with rel- 
atively tall grass and herbs, and with poor 
development of Ipomoea (Fig. 5, left) : 
Bi: Floor stratum corresponding to Ai. 
B 2 : Stratum of short grass and herbs. 
B 3 : Stratum of tali grass and herbs. 
C: Areas with short grass and herbs alone; 
Ipomoea cover is more developed than in 
B (Fig. 5, middle): 
Ci: Floor stratum corresponding to Bi. 
Co: Stratum of grass and herbs. 
D: Areas largely exposed, with patchy de- 
velopment of grass and herbs; Ipomoea 
cover less developed than in C: 
C' (Ci and C 2 ): Littoral zones corre- 
D' sponding to C and D 
in habitat structure. 
However, C'i con- 
sisted of scattered 
establishments of Ipo- 
moea frontiers alone, 
and D' is almost aphy- 
tic. 
H: Areas disarranged by human activities. 
The relative size of these habitats was ap- 
proximately A greater than or equal to C ap- 
proximately equal to D / > D > B approximately 
equal to C. The richness of each habitat in num- 
ber of species and in ecological endemicity may 
be roughly estimated by comparing the total 
species number with the number of species 
found exclusively in each habitat (see Table 4). 
Conclusions derived from these data are: 
1. With respect to vertical distribution, the 
floor strata are far richer both in species number 
and in ecological endemicity than are the upper 
strata. Apparently, this is caused by the poor 
development of vegetation in the latter. 
2. Horizontally, A is the richest section in 
