138 
Psyche 
[Vol. 94 
(8 individuals, 3%) reached pupation at the “field” site while larvae 
from 15 of 22 broods (96 individuals, 1 1%) reached pupation at the 
“forest” site (p <0.01, Fisher’s Exact test). 
Larval mortality in undefended groups 
The effect of maternal guarding on larval survival was investi- 
gated during August 1980 by removing mothers from some larval 
groups and comparing their survival to that of nearby guarded 
groups at similar stages of development. Initially, six nearest- 
neighbor pairs, groups with and without mothers, were found, 
marked and revisited daily. The unguarded group disappeared first 
in each of the six pairs. All unguarded larvae were gone within three 
days. One week later the procedure was repeated with another six 
pairs. The motherless group disappeared first in four pairs. Both 
groups of a fifth pair disappeared between visits and neither group 
of a sixth pair had disappeared at the end of the sixth day. General- 
ist predators, chiefly myrmecine ants and polybiine wasps, preyed 
upon some motherless larval groups within an hour after removing 
the mother. On one occasion a Polybia sp. wasp discovered a group 
within 5-10 minutes of removing the mother. The wasp returned at 
2-4 minute intervals to remove larvae until the group was com- 
pletely gone. 
Pupation. 
After consuming their last leaf, penultimate instar larvae moved 
1-2 m down their host plant to pupate, followed by their mother. 
The location of a pupal aggregation is often revealed by a stem 
which has been girdled by larvae as they descended to pupate. At the 
pupation site, often only 0.5 m or so above the ground, larvae form 
a tight cluster of bodies completely encircling the stem (Fig. lc). The 
mother normally stands motionless for long periods of time on the 
backs of her pharate larvae or pupae. Movement in the vicinity of 
the aggregation often stimulated her to begin walking briskly and 
repeatedly around and over the group. No interactions were 
observed between mothers on pupae and predators. 
Twenty-seven pupal masses were collected during September and 
October of 1978 from the “forest” and “field” sites and allowed to 
develop in the laboratory. Ten to 17 days passed in the lab before 
adults began to eclose. Teneral adults eclosed from 68 percent of the 
pupae, parasitoids from 14.3 percent and the remaining 17 percent 
