Ecology of Manana Island, Hawaii — Tomich et al. 
359 
the occasional fresh deposit, had a roughened 
and weathered appearance.. These phenomena 
were explained when we observed at night that 
the cosmopolitan terrestrial isopod, Porcellio 
laevis Latreille, swarmed over the rabbit pellets 
and fed on them. Often a single pellet had 
several Porcellio on it and was rolled about as 
the isopods shifted position. Thus the destruc- 
tion of rabbit feces is hastened and the usual 
methods of censusing rabbits by pellet counts 
cannot be applied to the Manana population. 
Three male body weights were 2189, 1695, 
and 1670 g and three females weighed 2055, 
1970, and 1640 g, indicating a similar weight 
in the two sexes. Average weight for the six 
rabbits was 1870 g. Paunched weight averaged 
73.6% of the live weight. A male and female 
collected in August 1961 weighed 2284 and 
2087 g, respectively. Paunched weight averaged 
70.4% as heavy. Both were fat, rating at 3 and 
2, respectively, on a scale of 0 (none) through 
3 (heavy). The rabbits observed in 1964 were 
in good flesh but not fat. On the same scale a 
pregnant doe registered 1, and four others 0 
for an average of 0.2. Five males were 1 and 
one was 0, for an average fatness of 0.8. 
The population had just passed a peak of 
reproduction. All five adult does were lactating, 
averaging 2.2 on a scale of 0 through 3. One 
doe was about 10 days pregnant and showed the 
least mammary tissue (rated at 1). Corpora 
lutea were counted easily in the pregnant doe 
as 8R-1L. Eight embryos were implanted in 
the right uterine horn, none in the left. Another 
doe had distinct black uterine scars 4R-0L and 
corpora lutea 7R-1L, suggesting a 50% pre- 
implantation or early resorption loss. No other 
scars were found. Uteri of the four parous does 
definitely were inactive as flabby flattened struc- 
tures 5 to 7 mm wide. The reproductive rate 
was seemingly high. These data on corpora 
lutea, embryos, and placental scars show that 
litter size may average about 6.0 (allowing for 
a 25% loss before birth of shed ova and early - 
stage embryos). 
The testes of one adult male weighed 2.7 g 
and the cauda epididymides had visible tubules 
indicating the presence of sperm. In another 
buck testes were larger than this, in two they 
were of similar size, and in two they were 
smaller. At least four and perhaps all of the 
six males were in breeding condition. Regres- 
sion may have been in progress. That the four 
parous does were sexually inactive suggests that 
the breeding season was nearly terminated. 
Each of the party who crossed the crater re- 
ported seeing small young rabbits. The smaller 
of two juvenile females collected weighed 352 
g and was still nursing, for it had a small mass 
of curds in the stomach. 
Age structure offers some clues about the 
population. Of 12 rabbits classified by age, 5 
were less than 10 months old (2 were ju- 
veniles), 5 were from 15 to 26 months, and 
2 were 33 to 38 months. The criteria used 
w r ere those of Taylor (1959) which employ 
the progressive closure of skeletal epiphyseal 
sutures. These three age groups are spaced at 
approximately yearly intervals and imply that 
breeding may normally occur yearly in a season 
extending from November to April. A rabbit 
attains adult size at the age of 4 to 5 months. 
Because the rabbits on Manana must depend 
on the few foods available, it is almost manda- 
tory that they eat all palatable plants within 
reach. The rabbits taken in 1964 all had full 
intestinal tracts. Reingestion, as described by 
Meyers (1955) and by Rowley (1956), was 
evidenced by soft pellets which made up 10% 
to 40% of the stomach contents of at least 
seven rabbits. None had other than hard rectal 
pellets, showing that reingestion takes place 
earlier in the day than the 4:30 to 6:00 pm 
period in which the rabbits were shot. 
Thorough mastication and the second pas- 
sage of food through the alimentary canal 
usually makes identification of food items a 
tedious process requiring specialized techniques. 
We were fortunate to have collected a male 
(2759) 4 which had lost its fourth and fifth 
upper cheek teeth on one side, resulting in 
malocclusion and uneven wear of all other 
cheek teeth. He swallowed most of his foods 
whole, allowing their easy identification (Table 
2). Comparative data for the other rabbits are 
given in Table 3. 
Grass fragments composed over half the 
identifiable materials in all seven animals, al- 
though volumetric measurements were not 
4 All mammal specimen numbers refer to the 
catalog of P. Q. Tomich. 
