46 
Psyche 
[Vol. 88 
Table 1. Three-way analysis of variance comparisons of the positions of larvae and 
pupae in 26 nests of P. montanus collected in 1978 and 1979, 3 nests of P. subnitidus 
collected in 1979, and 9 nests of P. rugosus collected in 1979. (As the data were 
expressed as percentages of the total nest population, they were subjected to an arcsin 
transformation before analysis.) 
Source 
df 
MS 
F 
P. montanus 
Different nests 
25 
0.005 
0.556 ns 
Positions of larvae and pupae 
1 
0.007 
0.778 ns 
Levels in nests 
7 
0.315 
35.000*** 
Nests X brood 
25 
0.004 
0.444 ns 
Nests X levels 
175 
0.110 
12.222*** 
Brood X levels 
7 
0.028 
3.111** 
error 
174 
0.009 
P. subnitidus 
Different nests 
2 
0.000 
0.000 ns 
Positions of larvae and pupae 
1 
0.000 
0.000 ns 
Levels in nests 
22 
0.017 
5.667*** 
Nests X brood 
2 
0.000 
0.000 ns 
Nests X levels 
44 
0.017 
5.667*** 
Brood X levels 
22 
0.003 
1.000 ns 
error 
43 
0.003 
P. rugosus 
Different nests 
8 
0.000 
0.000 ns 
Positions of larvae and pupae 
1 
0.000 
0.000 ns 
Levels in nests 
39 
0.019 
9.500*** 
Nests X brood 
8 
0.000 
0.000 ns 
Nests X levels 
312 
0.010 
5.000*** 
Brood X levels 
39 
0.003 
1.500 ns 
error 
311 
0.002 
larvae and pupae are placed in separate levels of the nests in any of 
the three species (Table 1). There is a significant difference between 
the levels of the nests, which is evident in Figures 6, 7, and 8. The 
brood tend to be in the lower levels of the nest. Although it is 
commonly assumed there is segregation of the larvae and pupae, 
statistical analysis has not been performed in the past to support the 
assumption. 
In one instance, a P. montanus nest placed a large number of 
brood on the soil surface near the nest entrance after a late-summer 
