Taxonomy of Rhinoloplius simplex 
7 
sample size variation. Cluster analysis used the 
UPGMA method of Sneath and Sokal (1973). All 
statistical analyses were undertaken with Genstat 5 
(Genstat 5 Committee 1987) and Biosys-1 
(Swofford and Selander 1989). 
MORPHOLOGY: STATISTICS 
Univariate statistics 
Mean, standard deviation, minimum and 
maximum values and sample size of each taxon 
are presented in Table 1 for (a) skull, dentary and 
dental characters and (b) external characters 
Multiple regressions 
Multiple regressions were run for skull, dentary 
and dental characters and external characters on 
sex and adult age for five taxa (R. borneensis, R. 
megaphyllus, R. simplex simplex, R. s. parvus and R. s. 
subsp. nov.). Additionally, multiple regression was 
run for the R. simplex group alone for skull and 
external characters on sex, adult age and island. In 
these analyses, islands with small sample size were 
omitted because either a sex or adult age category 
was absent. The results of these analyses are 
presented in Tables 2 and 3. 
In the following discussions, because of the large 
number of interactions being tested, the level of 
significance was set at Pc.Ol. 
All 5 taxa - skulls 
Sex. From Table 2a, eight characters (greatest 
skull length, GSL; braincase breadth, BB; rostrum 
height, RH; rostrum length, RL; outer cochlear 
width, CCW; upper maxillary tooth row length, 
C'M’L; lower tooth row length, I'M 3 L and dentary 
length, DL) showed a significant relationship with 
sex alone (PcO.OOO - P=0.002). Also there was a 
significant interaction for outer M 3 M 3 width for sex, 
age and taxon (P=.009). This interaction resulted 
from subadult females being smaller than subadult 
males in R. megaphyllus and females being slightly 
smaller than males in R. s. parvus and R. borneensis. 
Age. Rostrum height, RH; showed a significant 
(P=0.003) relationship with age alone. Outer M 3 M 3 
width also had a significant interaction between 
age, sex, and taxon (P=0.009), as discussed above. 
Taxon. All skull characters, except M 3 width, had 
a significant relationship (P=<0.001) with taxon. 
These relationships were consistent between the 
sex and age groupings; the only significant 
interaction was between sex, age and taxon 
(P=0.009) for outer M’M 1 width, as discussed 
above, clearly there is a considerable extent of 
morphological distinctness among these taxa. 
All 5 taxa - externals 
Sex and age. There were significant relationships 
with snout to vent length, SV, both with sex alone 
(PcO.OOl) and age alone (P<0.007) and with 
interactions between sex and age and sex and 
taxon (Table 2b). 
Taxon. All characters were significantly related to 
taxon (P<0.001), except basal sella length (BSL). 
R. simplex group - skulls 
Sex. No character had a significant relationship 
with sex alone, although there was a significant 
interaction between age, sex and island (P=0.005 
and P=0.002, respectively) for braincase breadth BB 
and M 2 width, M 2 W. For BB, this resulted from 
young adult females being larger than old adult 
females on Roti I., and males being larger than 
females on Moyo and Flores, whereas on other 
islands they were approximately the same size, For 
M 2 W, young adult males were larger than old adult 
males on Flores and young adult males were larger 
then young adult females on Savu I., whereas on 
other islands they were approximately the same 
size. Also, for P 1 width, P'W, there was a significant 
interaction between sex and island (P=0.008) which 
resulted from females being larger than males on 
Savu I., whereas on other islands they were 
approximately the same size. 
Age. No character had a significant relationship 
with age alone. But there were significant 
interactions for BB and M 2 W between sex, age and 
island, as discussed above. 
Island. All 30 characters had a significant 
relationship with island alone, most at P«0.001. 
These relationships were consistent for sex and age 
categories except for braincase breadth, M 2 width, 
and P’ width which had interactions between age 
and or sex and island, as discussed above. 
Clearly, there was marked morphological 
differences among the island populations of R. 
simplex. 
R. simplex group - externals 
Sex and age. The only significant relationship was 
the interaction between sex and island for ear 
length, EL (Table 3b). This resulted from the 
Rotinese sample where males had greater ear 
lengths than females, whereas on other islands they 
were subequal. 
Island. All characters except ear length, EL; 
maximum anterior noseleaf breadth, ALB, and 
maximum basal sella breadth, BSB, were 
significant, most at P<0.001. The only significant 
interaction was again between sex and island for 
ear length (Table 3b) as discussed above. 
Multivariate analyses 
5 taxa analysis 
Canonical variate (discriminant) analysis (DFA) 
was carried out on five taxa (R. megaphyllus, R. 
