Malaria and Rome: A History of Malaria in Ancient Italy Page 13
⁹⁸ Ginanni (1774: 133–4 n. 3).
⁹⁹ Interpretation of the situation is complicated by the fact that the Po delta region was an area of Greek colonization. It can be difficult to disentangle evolution in situ from gene flow caused by human migrations (the same problem occurs in Sardinia, discussed in Ch. 4. 3
Ecology of malaria
83
of b-thalassaemia mutations in this region closely matches the distribution of mosquito breeding habitats.¹⁰⁰ Thus Ginanni was right to argue that the local people were less likely to die in the short term from individual malarial infections. However malaria had profound long-term effects on such populations, sharply reducing overall life expectancy and distorting the age-structure of the population. These long-term effects were already obvious to Doni in the seventeenth century and will be described in more detail later on (see Chs. 5 and 11 below). Ginanni’s observations on the resistance of the local population to malaria are in fact paralleled by the situation in human populations in parts of tropical Africa today where malaria is endemic. In such African populations one is unlikely to die directly from malaria once one survives past a certain age in childhood—the precise age varies from region to region in Africa and depends on the transmission rate of malaria, in other words the frequency of infections—by which time acquired immunity has developed, but malaria nevertheless has a deep influence on age-specific life expectancy. Under such circumstances it is quite easy for both the inhabitants of regions with intense malaria themselves and outside observers to jump to the conclusion that malaria is not a major problem. The only legitimate inference to be drawn from such a conclusion, as in Ginanni’s case, is that it demonstrates a lack of understanding of the immunology of malaria—not surprising in an eighteenth-century author. It does not constitute evidence for any mildness or avirulence of the disease. The studies that were done at the beginning of the twentieth century showed that P. falciparum was common in the coastal regions of northeastern Italy alongside P. vivax and P. malariae, although P. vivax lasted longer than P. falciparum once eradication measures commenced. A survey in 1849
below). Greeks were present in the cities of Atria and Spina from the sixth century
onwards (intermingled with Etruscans, other local Italic peoples, and later on Celts as well), who might well have carried with them genetic mutations related to malaria from their homeland in Greece, another area of endemic malaria. Consequently thalassaemia, for example, may already have occurred in northeastern Italy in the fifth century even though the area did not yet suffer from malaria.
¹⁰⁰ In the province of Ferrara allele frequencies for b-thalassaemia mutations decrease from the Adriatic coast westwards, while in the province of Rovigo there is no such decrease (Barrai et al. (1984) ). In Ferrara, south of the river Po, marshy areas were largely confined to the east of the province, while in Rovigo, north of the Po, there were marshy areas throughout the entire province.
84
Ecology of malaria
14. Anopheles sacharovi, a very important vector of malaria in Italy and eastern Mediterranean countries. © The Natural History Museum, London.
concluded that 16.7% of the entire population of the province of Venezia suffered from malaria in that particular year.¹⁰¹
In recent times Anopheles sacharovi was common in certain foci of intense malaria around Ravenna, breeding in brackish water, although A. labranchiae was virtually absent from this region. Presumably in antiquity the hydrological conditions of the marshes around Ravenna were unsuitable for A. sacharovi, preventing this species, the most efficient vector of malaria in Mediterranean countries, from breeding in the area. Gilberto Corbellini and Lorenza Merzagora observed that the early Venetian writers on mal’aria (see Ch. 2 above) emphasized the importance of the mixing of fresh water and seawater for the generation of malaria. This shows that the modern ecology of the coastal regions of northeastern Italy, with brackish conditions favourable for A. sacharovi, already existed by the end of the medieval period. Among the more zoophilic species of Anopheles mosquito, A. maculipennis s.s.
prevailed to the north of Venice, while A. atroparvus (a secondary ¹⁰¹ Rausa and Romano (2000) and Hirsch (1883: 213–14) described the distribution of malaria in northern Italy in the nineteenth century.
Ecology of malaria
85
malaria vector in northern Italy), was common in the Po delta, and A. messeae also occurred in places. There certainly were plenty of mosquitoes in northeastern Italy in antiquity, since Martial mentioned the culices of Atria, another town in the same region, while Sidonius Apollinaris mentioned the mosquitoes of Ravenna. However, we may infer that these were predominantly not anthropophilic anopheline species, but zoophilic anopheline species or culicine mosquitoes. Culicine mosquitoes can breed in dirty water but do not transmit malaria to humans, although they are the vectors of the avian species of malaria.¹⁰² The Po delta, like other major Mediterranean deltas such as the Ebro in Spain and the Nile in Egypt, has increased in size substantially within the last three or four centuries, as a result of very extensive alluviation following that which had already taken place in antiquity. This process of alluviation, a very important theme to return to later, led to the abandonment of a number of towns which functioned as ports in various parts of the Mediterranean. For example, Strabo observed that Spina in the Po delta region, which was on the coast when it was founded, was ninety stadia inland by the first century . Similarly Luni on the other side of Italy was stranded in the same way, and malaria moved in, just as it did at Ravenna. Further south in Campania, the Greek colony of Cumae had a lagoon on its northern side connected to the sea, which provided a sheltered harbour when the colony was founded in the eighth century . However, by the end of the first century the lagoon entrance had become silted up, destroying the harbour. Subsequently the lagoon started to fill in slowly from the north, until it was finally completely drained by Mussolini. These are all illustrations of a very widespread, long-running process of change. Virtually all the coastal regions of Italy were completely different in antiquity from the way they look today.¹⁰³ Thus the development of the coastal malarial environments of Italy as they appear today was a prolonged process which unfolded over a period of several thousand years.
In Lazio the delta of the River Tiber underwent a similar process ¹⁰² Martial 3.93; Sidonius Apollinaris, Letters, 1.8.2. to Candidianus. Neri and Gratch (1938) and Zamburlini (2000) discussed the mosquitoes of Ravenna and the Veneto. See also Christie and Gibson (1988); Veggiani (1973) and (1986: esp. 26) on malaria.
¹⁰³ Strabo 5.1.7.214C and 5.2.5.222C; Potter (1981: 7) on Spina; Laquidara (1989) on malaria at Luni; Vecchi et al. (2000) on Cumae. Spivey and Stoddart (1990: 24–6) described the changes in the coastline around Pisa and Populonia. Hunt et al. (1992) described one example of continuing erosion in Tuscany within the last millennium.
86
Ecology of malaria
of development. The history of the Roman port of Ostia, at the mouth of the river, in relation to malaria merits some discussion.
Ostia was probably founded to exploit salt deposits at the mouth of the Tiber.¹⁰⁴ It was probably not unhealthy to begin with, like Ravenna, because the salinity level was too high for Anopheles mosquitoes. Nevertheless there was always a potential for the spread of malaria in the vicinity because of Tiber floods. The opinion of Meiggs, namely that Ostia must have been healthy during the time of the Roman Empire simply because it was a flourishing town, is an untenable argument from silence.¹⁰⁵ It is very hard to believe that Ostia was perfectly healthy in the second century , at the time when Galen regarded malaria as endemic in the city of Rome itself (see Ch. 8 below), and there certainly are explicit signs that Ostia was at the very least becoming unhealthy in late antiquity. St. Augustine’s mother, Monica, died from malaria at Ostia, after an illness lasting nine days, on the way from Rome back to Africa, while Augustine himself al
so contracted malaria at Rome.
Justinian’s general Belisarius suffered from a life-threatening fever at Ostia in 536.¹⁰⁶ The cases of Belisarius, Augustine, and his mother are classic illustrations of the general principle that immigrants were more vulnerable than natives in areas where malaria was endemic. The spread of malaria at Ostia intensified in post-classical times, as sediment deposition created and enlarged an alluvial plain on both sides of the river near its mouth, steadily reducing the size of the large early Holocene lagoons behind the coastal barrier on either side of the river. Procopius noted the abandonment of Ostia in favour of Portus on the other side of the Tiber in late antiquity owing to silting of the harbour. Modern research in geomorphology has confirmed that the major extension of the ancient coastline seawards towards its current location has occurred principally since about 1500, when the modern marine-delta area began to move beyond the coastal barrier separating the ancient lagoons from the sea.¹⁰⁷ However, the letters of ¹⁰⁴ Grandazzi (1997: 74–83).
¹⁰⁵ Meiggs (1973: 101, 233).
¹⁰⁶ St. Augustine, Confessions 9.8, 11: decubuit febribus . . . die nono aegritudinis suae . . . anima illa religiosa et pia corpore soluta est (she developed fevers . . . on the ninth day of her illness . . . that religious and holy soul was released from her body); 5.9: ingravescentibus febribus (fevers becoming worse); Procopius, BG 3.19.
¹⁰⁷ LeGall (1953: 24–5, 320–2); Bellotti et al. (1995); Dionysius Hal. AR 3.44; Procopius BG
1.26.3–13.
Ecology of malaria
87
St. Peter Damian prove that there was endemic malaria in Ostia as in Rome by the middle of the eleventh century , when he wanted to relinquish the bishopric of Ostia because of the unhealthiness of Ostia (see Ch. 8 below). Consequently environmental changes favourable to the mosquito vectors of malaria had already occurred in the vicinity of Ostia before the major modern progradation of the Tiber delta. The spread of malaria at Ostia was associated with the filling in of the lagoons behind the ancient coastal barrier after they had become isolated from the sea. Early modern writers unanim-ously concurred that the area of the notorious Maccarese swamp north of Ostia was virtually uninhabitable.¹⁰⁸ This state of affairs may go back to antiquity in the light of Silius Italicus’ description of Fregenae, which can be identified with modern Maccarese.¹⁰⁹
Just like the lagoons behind the coastal barrier near Ostia, the Pontine Marshes in southern Latium were also cut off from the sea.
Pliny the Younger confirms Vitruvius’ description of the hydrology of these marshes and indicates that the water table was very high at his villa near Laurentum. Anopheles mosquitoes do not lay their eggs in seawater or water with a high salinity level, although Grassi carried out experiments which showed that their larvae can survive in a 2:1 mixture of fresh water: seawater. Similarly Celli noted that Anopheles larvae did occur in the weaker sulphurous waters of Sezze, a town whose problems with malaria have already been noted. However, they were not found in the strong sulphurous waters of Tivoli in Latium, the famous Aquae Albulae which were reputed in antiquity to be very good for healing wounds, although there must have been mosquito breeding sites somewhere in the neighbourhood, since malaria impeded the operation of the quarry which produced lapis tiburtina near Tivoli in the nineteenth ¹⁰⁸ Hare (1884: i. 40–1) wrote as follows: ‘the peasants do all their field labour here in gangs, men and women together . . . they have…the constantly recurring malaria to struggle against, borne up every night by the poisonous vapours of the marsh, which renders Ostia almost uninhabitable even to the natives in summer, and death to the stranger who attempts to pass the night there’, cf. Tomassetti (1910: ii. 496–507). Knight (1805: 100) wrote that ‘the air is particularly unhealthy, and the town is chiefly inhabited by galley slaves, who work in the salt mines’. Blewitt (1843: 524) wrote that at Ostia ‘during the summer heats . . . the neighbouring coast is severely afflicted with malaria’. Mammucari (1991: 156–7, 162–3) reproduced paintings of the Maccarese forests and marshes by Onorato Carlandi and Cesare Bertolla made before Mussolini’s bonifications.
¹⁰⁹ Silius Italicus 8.575: obsessae campo squalente Fregenae (Fregenae surrounded by a plain overgrown with weeds through neglect).
88
Ecology of malaria
century.¹¹⁰ The precise chemical composition of water is very important.
Hackett commented as follows on the complexity of biological phenomena: ‘for the most part the reasons for these secular oscillations in the curve of malaria endemicity are inscrutable, because the internal adjustment of the elements composing a biological situation in dynamic balance is inconceivably complicated’.¹¹¹ His words are still as true today as when they were written (see Ch. 9
below for discussion of the idea of ‘secular oscillations’). Nevertheless research over the last thirty years has made it absolutely certain that the coastal environments of Italy were not static, but were, rather, changing continuously from antiquity right up to the bonifications of modern times. Writing at the end of the last century, Celli was unable to conceive of major environmental changes in the vicinity of Rome and concluded that the periodic fluctuations which he thought he could discern in the severity of malaria in the Roman Campagna could only be explained in terms of variations in the virulence of P. falciparum malaria. However, now that it is known that the physical environment has been changing continuously, it is possible to entertain other explanations for any periodic variations in the degree of endemicity of malaria. It is essential to follow this approach because, as was noted in Ch. 3 above, the most recent scientific research indicates that extreme virulence is adaptive for P. falciparum. There is no reason for supposing, pace Celli, that its virulence ever diminished in historical times. Moreover, the frequency of malaria could also have been influenced in the past by changes in the balance between different mosquito species. Mario Coluzzi expressed to the author his view that there is intense competition between different species of mosquito and that the balance between different species of mosquito is constantly changing at the local level in Italy. Evidently there could have been countless changes in the balance between different species in the past as well.
Although the details remain obscure, and will probably always be obscure, it is inevitable that major alterations in the habitats available to mosquitoes for breeding purposes will have a huge ¹¹⁰ Pliny, Ep. 2.17.25; Celli (1900: 77, 80); F. Giordano Condizioni topografiche e fisiche, in Monografia (1881: xxxiv); Vitruvius 8.3.2; Pliny, NH 31.6.10; Seneca, QN 3.20.4; LeGall (1953: 51); Quilici (1979: 64–5).
¹¹¹ Hackett (1937: 4).
Ecology of malaria
89
impact on species which are so sensitive to environmental parameters. Mosquitoes themselves can sometimes modify their own patterns of behaviour to their own advantage. The contemporary example is the tiger mosquito, Aedes albopictus, the vector of the virus of dengue fever in its homeland in south-east Asia. This denizen of tropical rainforests has started to use rubber tyres as a breeding site.
It is thought to have been transported around the world in this way and is now spreading in Mediterranean countries, even though it does not fly further than about two hundred metres from its breeding sites. It was first observed in Italy at Genoa in 1990, and has since spread to Campania, Lazio, Tuscany, Veneto and other regions.¹¹²
Today the various species belonging to the Anopheles maculipennis complex prefer fresh water. However, the three anthropophilic species which are dangerous vectors of human malaria in Europe ( A. labranchiae, A. sacharovi, and A. atroparvus) have a greater tolerance of slightly brackish conditions than the other, more zoophilic species. A. sacharovi can tolerate higher salinity levels than the other two vector species. Lancisi in the eighteenth century reckoned that coastal marshes which sometimes received seawater were particularly dangerous (e.g. at Ostia, Ferrara, and Ravenna).¹¹³ Consequently the anthropophilic species have
an adaptive advantage over the zoophilic species in coastal environments, so long as the degree of salinity of the water is not too high. This explains why malaria in Italy tended to have a predominantly coastal distribution. The dangerous anthropophilic species certainly also occurred inland, especially in salty regions such as Diamantina in the Po valley, but faced greater competition there for suitable breeding sites from the harmless zoophilic species.¹¹⁴
¹¹² Sabatini et al. (1990).
¹¹³ Missiroli (1938: 14); Lancisi (1717: 21). Sambon reported finding larvae of what was then called A. maculipennis very plentiful in brackish water at Ostia, ‘amongst large brown clumps of floating algae and seaweed’ (Sambon 1901 a: 198).
¹¹⁴ Hackett (1937: 26, 60, 106). Koella et al. (1998) produced some evidence that malaria parasites can alter the feeding behaviour of mosquitoes by inducing them to bite humans more frequently, another illustration of the complexity of the phenomena under consideration.
90
Ecology of malaria
4. 3 M S
On Sardinia A. labranchiae exploits a lack of competition to breed in environments, such as mountain streams above 500–1000 metres, where it is rare on the mainland of Italy, since several other species of mosquito ( A. atroparvus, A. claviger, A. maculipennis s.s. , A. subalpinus, A. superpictus) which are common on the mainland of Italy are either rare or absent from Sardinia.¹¹⁵ This helps to explain the greater intensity of malaria on Sardinia compared to the mainland of Italy in the past. Since the intensity of malaria on Sardinia was already obvious to Roman authors two thousand years ago, it may be inferred that the distribution of mosquito species on Sardinia at that time already matched the modern distribution patterns.¹¹⁶