1984] 
Shapiro — Pieris rapae 
165 
All larvae were placed in outdoor mesh cages over fresh cuttings 
of their hosts, and allowed to complete development under ambient 
conditions. The cages were less than 0.3 km from the U.C. Davis 
campus meteorological station. The first pupa of P. zelicaon was 
formed 13 January and the last 24 February. 50 of the 67 rapae 
larvae were parasitized by Apanteles (Cotesia) glomeratus (L.) 
(Hymenoptera: Braconidae) and failed to pupate; 5 died of un- 
known causes; 12 pupated between 9 January and 28 February. 
Results and Discussion 
Of the nine winter zelicaon, 1 prepupa died of unknown causes; 5 
eclosed without diapause; and 3 apparently entered diapause. 
Eclosions occurred on 21 and 24 February and 1, 5, and 12 March 
1984. As can be seen from Table 1, these dates coincide with the 
flight of P. zelicaon in the field. (At Gates Canyon, Vaca Hills, one 
zelicaon was found as early as 19 February 1984, the earliest flight 
ever recorded in the region.) All the butterflies which emerged were 
female, as was one of the 3 diapausers. 
Of the 12 healthy rapae pupae, all eclosed between 9 February 
and 24 March (seven males, five females). The flight began some- 
what earlier afield but continued throughout this period. 
The Apanteles wasps emerged without diapause in synchronized 
batches, each from its own host, between 24 January and 11 
March — again, well-timed to parasitize first-instar larvae from eggs 
laid by first-brood rapae females. Diapause in this species is known 
to be under direct photoperiodic control (rather than mediated 
through the host, hormonally) in the U.S.S.R., but has not been 
studied in the introduced North American populations (Danilevskii, 
1961). 
The phenotypes of emerging adults of both butterflies were 
compared with long series of field-collected specimens from Suisun 
and Davis from 1984 and prior years, including many individuals 
which must have come from diapaused pupae, and with reared ex- 
diapause individuals. No phenotypic differences which might permit 
the detection of non-diapaused adults in the spring populations 
were recognized. This result confirms experimental results which 
indicated that the post-diapause adult phenotype is not physio- 
logically coupled to the prior developmental arrest, but is rather a 
function of ambient conditions after reactivation (Shapiro, 1975a, 
1978), superimposed on irreversible short-day prediapause deter- 
mination. 
