parasite had been tricked into allocating more of its gametocytes into
males at a time when the male gametes were able to search for females
unimpeded and that therefore there was a shortage of females. Thus sex
allocation appears to be adaptive and is able to compensate for the
decreasing efficacy of the male gametes (due to host immunity) by pro-
ducing more male ‘insurance’ gametocytes in response to erythropoietin,
induced by the anaemia provoked by the asexual parasite proliferation.
However, in human and lizard malaria parasites, there seems to be no
clear correlation between sex ratio and transmission success (James, 1931;
Schall, 2000). Although there may be considerable differences between
malaria species, most notably the host immunological response to infec-
tion, further consideration of parasite life histories, with special reference
to the human malariaP. falciparum, may elucidate this apparent paradox.
In regions of endemic malaria, where most of the population has
some immunity, the majority of sporozoite inoculations result in a
low asymptomatic parasitaemia, which may last for several months
(Greenwood, 1987). Such chronic infections are significant for parasite
persistence and the infectious reservoir (Jeffery and Eyles, 1955), espe-
cially in regions with a very short transmission season, where parasites
must survive for up to 10 months without transmission (Arnot, 1998). It
is assumed that parasites restart transmission upon the arrival of the
mosquitoes but must achieve fertilization with a very low gametocyte
density, which demands an extremely efficient fertilization process,
opposite to that seen during the acute phase of infection. As can be
deduced from a simple mechanistic model of fertilization, one way to
ensure fertilization success when increasing gametocyte density is not an
option would be to increase the proportion of males (Paulet al., 1999a).
Hence a clone can find itself in hosts of widely ranging quality (immune
response) or in a condition (high, acute vs. low, chronic density) deter-
mining its reproductive capacity. In such ‘low-quality’ conditions, mating
may no longer be assured. Malaria-parasite clones will often find them-
selves alone in a host and, because sexual reproduction is a prerequisite of
transmission, clones must therefore self-fertilize to ensure transmission
success. This selects not only for the maintenance of a hermaphroditic
state, but also for an ability to adjust its male : female function investment
to ensure fertilization as host quality varies. Thus, although parasite
population structure (number of parasite clones per person) may promote
sex allocation under LMC, variable host quality, which influences
asexual parasite density (i.e. foundress condition) and hence gametocyte
numbers, can also significantly affect mating assurance and so influence
the optimal sex-allocation strategy.Concluding Remarks
Parasites must exist in discrete habitats of highly variable quality,
between which they must transmit. In addition to developing strategies toParasite Sex Determination 215