females actively pollinate yucca flowers after ovipositing and thereby
ensure a food supply for their seed-eating progeny. A flower that receives
a high number of eggs is more likely to be selectively aborted, which
results in the death of all eggs (Huth and Pellmyr, 1999) (differential
abortion may be a means by which the plant ensures that only a fraction of
its developing seeds is consumed). Thus, the plant’s response will itself
prevent competition for food among moth larvae, but it still ‘pays’ for a
female not to overexploit a host (and risk destruction of her eggs).
Yucca moth females do lay fewer eggs on previously visited hosts,
which they appear to recognize through a marking pheromone (Wilson
and Addicott, 1998; Huth and Pellmyr, 1999). Some evidence suggests
that females can even assess the density of conspecific eggs quantitatively,
as noted before in the seed beetleC. maculatus. The yucca moth–yucca
interaction is unusual because the progeny of early-arriving females can-
not usurp resources from the progeny of late-arriving females; both early
and late eggs will die as a consequence of floral abortion. Huth and
Pellmyr (1999) have argued that these conditions favour effective host
marking by the first female to deposit eggs. In seed-parasite/pollinator
systems without selective abortion of flowers, plants use other means
to protect seeds (Jaegeret al., 2000, 2001), and host discrimination will
probably reflect the usual avoidance of competition or cannibalism.Variable Host Discrimination by Seed Beetles
Predictions from foraging models have rarely been tested with seed
parasites; most examples of adaptive variation in host discrimination
have involved parasitoids. Experiments using Callosobruchus seed
beetles confirm the importance of the larval competition curve in deter-
mining the degree of host discrimination. These beetles also provide
examples of adaptive plasticity in foraging behaviour as a result of a
female’s physiological state or experience.
Interfertile geographical populations ofC. maculatusshow striking
differences in the mode and intensity of larval competition within
seeds (Toquenaga, 1993; Messina, 1998; and references therein). In
some populations, larvae engage in a strong ‘contest’ type of competition
(including biting behaviour), so that small seeds virtually never yield two
adults and even large seeds yield few adults per seed. Larvae from other
populations display more of a scramble type of competition; larvae form
burrows near the periphery of the seed and burrows are less likely to inter-
sect. In these populations, small seeds often yield two adults and more
than ten adults can emerge from a large seed, albeit with reduced size.
Hybridization experiments revealed additive inheritance of differences in
larval competitiveness within seeds (Messina, 1991b).
Variation in larval competitiveness produces very different larval
competition curves, such as those depicted in Fig. 4.2. In populations
with strongly competitive larvae, the single-host maximum is at most a74 F.J. Messina