256 robert sallares
populations not affected by malaria (Sallares, 2002: 151–167). The impact of malaria
at the population level in the Mediterranean in the past can be recognized in two dif-
ferent ways. First, malaria was the only major infectious disease known to have existed
in the Mediterranean in the past to have an epidemiological association with mosquito
breeding sites in wetlands, since its two most important vectors are happy to breed in
marshes. (Mosquito-borne diseases in other parts of the world are transmitted by
other species of mosquitoes with different habits.) Consequently there is no doubt
that descriptions of wetland environments that were very unhealthy in summer, bear-
ing in mind that malaria is a temperature-dependent disease as may be found for
example in the ancient Hippocratic text Airs, waters, places, dating to the fifth century
bce or in literature from the ancient to the early modern periods describing the noto-
rious Pontine Marshes (Ilvento, 1934) south of Rome in central Italy, do indeed refer
to endemic malaria.
Second, studies of the value of clinical symptoms for malaria diagnosis in Africa do
suggest that during the malaria season (the wet season in tropical Africa today, the
summer in Europe in the past), in areas where malaria is known to be common, the
major clinical symptoms (alternating fevers and chills, sweating, headache, enlarged
spleen) can be used to make a diagnosis which in most cases will be confirmed by
subsequent laboratory analysis (Sallares, 2005: 208). Consequently, although there
will always be some uncertainty about retrospective diagnosis in individual cases, it is
not difficult to demonstrate the prevalence of malaria at the population level solely on
the basis of the sort of symptoms that are mentioned in historical accounts. This was
particularly the case in Europe, where malaria frequently affected adults because of low
transmission rates (diagnosis on the basis of clinical symptoms alone is more difficult
in children than in adults). Thus when, for example, St Willibald developed alternat-
ing fevers and chills on a pilgrimage from England to Rome, a known malaria zone,
in 722 ce, the overwhelming balance of probability is that the saint had been infected
with malaria, although it is impossible to exclude totally other possibilities (McCormick,
2001: 131). It should also be pointed out once again that the evidence of historical
texts for the Mediterranean pathocoenosis will increasingly be supplemented in the
future by data from biomolecular archaeology (for example, Sallares et al., 2004 for
malaria in Umbria in central Italy in the fifth century ad; Lalremruata et al., 2013 for
malaria in the Fayum in ancient Egypt) as well as data from conventional palaeopa-
thology (Sabbatani, 2008 and 2009).
Establishing that falciparum malaria dominated the Mediterranean pathocoenosis
in the past still leaves open the question of what was actually distinctive about the
Mediterranean pathocoenosis. After all, malaria occurs in many other parts of the
world today, as well as in northern Europe in the past. The clear answer to this ques-
tion is that differences in the habits of different mosquito vector species in different
areas, along with climatic differences and differences in agricultural systems, did com-
bine to create an epidemiology of malaria in the Mediterranean world that was sub-
stantially different from its epidemiologies in both tropical Africa and northern
Europe. In Africa south of the Sahara the most important vector species, Anopheles
gambiae, first, does not breed in wetlands (unlike European vector species), second, is
active all year round in areas with a suitable climate (unlike Europe, where only sea-
sonal transmission is possible because of malaria’s temperature requirements), and,
third, is a much more efficient vector of malaria. These factors combine to create an