Malaria and Rome: A History of Malaria in Ancient Italy

by Robert Sallares

  Other literary sources from late antiquity show that marshy environments were still feared then. Palladius duly made the standard observations that marshy and pestilential land should be avoided.⁶⁰ He then added a more interesting comment:

  In any event a marsh must be avoided, especially if it faces south or west and tends to dry up during the summer, because of the pestilence or hostile little animals which it generates.⁶¹

  Palladius’ observation of the extreme peril posed by marshes which desiccate during the summer leads on to consideration of the exact requirements of Anopheles mosquitoes for breeding pur-

  ⁵⁸ McFalls (1984: 118); Zei et al. (1990); Dobson (1997: 340–2); Barbosa and Arjona (1935: 11–18); Torpin (1941); Diagne et al. (2000) showed that the period of increased susceptibility to malaria continues for about two months after birth.

  ⁵⁹ Sallares and Gomzi (2001); Sallares et al. (2002); North (1896: 24) observed that the vicinity of Orte is particularly vulnerable to Tiber floods.

  ⁶⁰ Palladius, opus agriculturae 1.5.5, ed. Rodgers (1975): situs vero terrarum neque planus, ut stagnet; 1.5.6: qui ager pestiferi more fugiendus est.

  ⁶¹ Palladius 1.7.4: palus tamen omni modo vitanda est, praecipue quae ab Austro est vel occidente et siccari consuevit aestate, propter pestilentiam vel animalia inimica quae generat.

  Ecology of malaria

  69

  poses. Their larvae generally require water that is clear, in contrast to the larvae of Culex mosquitoes (non-transmitters of human malaria) which are happy in dirty water, although Anopheles larvae were also found in puddles of dirty water at Grosseto in the nineteenth century.⁶² In addition, the water must be well oxygenated.

  This explains the success of spreading oil on the surface of pools as a modern control measure. Anopheles larvae prefer pool and canal margins where there is plenty of vegetation, which provides cover from predators. Sambon noted that at Ostia in the summer of 1900

  all the pools and canals had enormous numbers of frogs, several species of fish (especially the young of the grey mullet), and swarms of dragonfly larvae and water beetles.⁶³ Consequently large permanent lakes are not necessarily any better for mosquitoes for breeding purposes than very small pools, which are only filled with water seasonally. Strabo noted that the large inland lakes of Tuscany, such as Vico, Bolsena, Chiusi, and Bracciano, produced substantial quantities of fish for the city of Rome.⁶⁴ A series of lakes along the coasts of Tuscany and Latium, such as the Prilius lacus near Grosseto and the lakes near Circeii, were also extensively exploited for fishing in Roman times.⁶⁵ In addition, the Romans created numerous artificial fishponds ( piscinae) attached to coastal villas.

  Such fishponds have been found at many locations along the coast of Etruria, such as Cosa, Torre Valdaliga, Grottacce, Pyrgi, and also in Latium and Campania. Recent research into these artificial fishponds concludes that their water was usually brackish: The archaeological evidence is decisive in showing that the Romans preferred brackish conditions for their seaside piscinae . . . Arrangements for mixing salt water and fresh water to create a brackish environment were ⁶² Celli (1900: 79) mentions Grassi’s observations at Grosseto. Austen (1901) discussed the differences between Anopheles and Culex mosquitoes.

  ⁶³ Sambon (1901 a: 199).

  ⁶⁴ Strabo, 5.2.9.226C. Quilici (1979: 104–6), noting the frequent presence of fish in offerings in archaic tombs in the Roman Forum, stressed that the River Tiber itself was an important source of fish in the early stages of Roman history, although it was wholly inadequate (and heavily polluted) by the time of the Roman Empire, cf. Nutton (2000 b: 66) and LeGall (1953: 267–8, 318–19). Pratesi and Tassi (1977: 43) described the rich fauna of fish in the modern Lago di Bolsena. Dennis (1878: 30) interpreted the prodigy in Livy 27.23.3, Volsiniis sanguine lacum manasse (the lake at Volsinii flowed with blood), as a sign that the area around the Lago di Bolsena was becoming unhealthy in the late third century . Magri (1999: 173) noted that the settlement of Monte Bisenzo nearby was finally abandoned in 1816 because of mal’aria arising from the Lagaccione marsh.

  ⁶⁵ On the Prilius lacus see Cicero, pro Milone 74, with Celuzza (1993: 92–3) for the identification of the site of Clodius’ villa; Pliny, NH 3.5.51.

  70

  Ecology of malaria

  common features in Roman fishponds and are well represented in the archaeological record from both the late Republic and early Empire.⁶⁶

  The chemical composition of water is very important for the breeding of Anopheles mosquitoes. Since they certainly can breed in brackish water, so long as it is not too salty, the water of some of these fishponds could have been favourable for malaria. The fish were an unfavourable factor for mosquito larvae, but a great deal would have depended on how much vegetation was present in these artificial fishponds. In any event, it is certain, as will be seen, that the coasts of Etruria and Latium were severely affected by malaria in the Late Republic and Early Empire. Strabo (quoted in Ch. 6 below) described the area around Circeii, which was a centre of fish farming, as pestilential. Consequently neither the hydraulic works of the fishponds nor their fish seriously impeded mosquito larvae.

  The vicinity of the large inland lakes mentioned by Strabo was certainly heavily infested with malaria in the early modern period.

  Nevertheless, Anopheles larvae can flourish at least as well in the small depressions in the ground, only containing water for a few months each year, that frequently occur on the lower slopes of hills in Latium and Tuscany, and particularly in the undulating terrain of the Roman Campagna. All the literature on the early modern Roman Campagna lays stress on the importance of small seasonal pools and puddles for the generation of malaria. Tommasi-Crudeli, scaling up the results of a small field survey to provide an estimate for the whole Roman Campagna, suggested that there were quite literally thousands of breeding sites for mosquitoes in the Roman Campagna in the nineteenth century, before many of the modern drainage operations. This was also the origin (besides rivers and streams) of much of the malaria in the past in southern Italy, where it frequently occurred in the absence of marshes or large lakes.⁶⁷ Consequently the most dangerous wetlands were those which tended to dry up in the summer, as Palladius observed, since there were no predators to eat the mosquito larvae. The ⁶⁶ Higginbotham (1997: 16); Rustico (1999) has also discussed the Roman fishponds recently.

  ⁶⁷ e.g. North (1896: 113–15) emphasized that pools which dried up in summer were very dangerous; Tommasi-Crudeli (1892: 34). Tommasi-Crudeli and North also discussed this phenomenon in the region around Mantua, showing that it was not confined to Lazio, cf.

  Dobson (1997) on England.

  Ecology of malaria

  71

  anonymous author of a late-eighteenth-century discourse on mal’aria noted that many of the localities in the Roman Campagna which were marshy in winter became completely dry in summer.

  Similarly in the marshlands of early modern England mortality from P. vivax malaria increased after dry summers as the marshes dried out.⁶⁸ The pseudo-Aristotelian Problems noted that after a wet spring and dry summer the autumn is lethal to all age-groups, but especially to children, the main victims of P. falciparum malaria when it is endemic.⁶⁹

  That even large lakes were subject to major fluctuations in central Italy in antiquity is shown by the example of the former Lago di Fucino, in the territory of the Marsi beyond the Monti Simbruini, which experienced great fluctuations in its water level according to Strabo. Julius Obsequens recorded that the Lago di Fucino overflowed its banks for five Roman miles in all directions in 137 .

  Draining it was another of Julius Caesar’s unfulfilled dreams. The emperor Claudius attempted at great expense to drain it. However, Tacitus’ account, stating that the tunnel that was originally constructed did not even reach half-way down the lake, illuminates the difficulties faced by the Romans in effectively performing major drainage operations. Claudius’ works were negle
cted by Nero.

  Later attempts in antiquity were made by Trajan and Hadrian, but the lake was not drained completely and permanently until the operations of Prince Torlonia commenced in  1862.⁷⁰ The plain ⁶⁸ Anon. (1793: 23); Dobson (1980: 378–80); see Ch. 5. 4 below.

  ⁶⁹ [Aristotle,] Problems 1.19.861b: Di¤ t≤, ƒ¤n toı ceim0noß bore≤ou ka≥ toı πaroß not≤ou ka≥ ƒpÎmbrou tÏ qvroß l≤an aÛcmhrÏn gvnhtai, qanat0deß g≤netai tÏ metÎpwron p$sin, m3lista d† to∏ß paid≤oiß, ka≥ to∏ß £lloiß d† dusenter≤ai ka≥ tetarta∏oi crÎnioi g≤nontai ƒn aÛt‘ (Why is it that if the north wind prevails in winter, and the spring is damp and rainy, and the summer is very dry, then the autumn is deadly to all, particularly children, while others suffer from dysentery and prolonged quartan fevers?). This is a typical description of a year ending with an autumn epidemic of malaria. The pseudo-Aristotelian Problems also contains a lot of other material on the seasonality of disease patterns.

  ⁷⁰ Strabo 5.3.13.240C. Leveau (1993) discussed the ideology of Roman attempts to manage the Fucine Lake. Pratesi and Tassi (1977: 222–5) described the modern environment of the basin. Giraudi (1989) suggested that the lake’s water level was low during the period c.300

  – 200, followed by the Roman drainage, which he unjustifiably assumed to have been a complete success. The words of Julius Obsequens for 136 , M Lacus Fucinus per milia passuum quinque quoquo versum inundavit—(the Fucine lake overflowed its banks by five miles), show that Giraudi’s generalizations must be taken cautiously. For the problems which beset the Roman drainage schemes see Tacitus, Annals 12.56–7; Pliny, NH 36.24.124; Suetonius, Claudius 20–21; Dio Cassius 61.33.5. Thornton and Thornton (1985) discussed the drainage works. The inscription CIL 9.3915 and SHA Hadrian 22.12 record the efforts of Trajan and Hadrian.

  72

  Ecology of malaria

  of the former Lago di Fucino might be too cold today for P. falciparum, situated as it is at an altitude of 669 metres (but not for P. vivax). However, the water of the lake exercised a moderating influence on temperature in the past. In antiquity the lake was once surrounded by a primeval forest, Angitia, mentioned by Virgil.⁷¹

  Deforestation of the mountain slopes encircling the basin, coupled with the effects of the modern drainage operation in 1862, significantly altered the microclimate of the region. It is now colder than it used to be, with more severe winter frosts. These climatic changes have eliminated the olive trees which formerly grew around the lake. They also gradually helped to convert the Fucine basin into an area of anophelism without malaria by the end of the nineteenth century. However, in the immediate aftermath of the drainage operations which commenced in 1862 it was observed that the frequency of malaria actually increased, presumably because the drainage left exposed areas of damp soil which had previously been permanently flooded, creating new breeding sites for mosquitoes. The example of the Fucine Lake illustrates the complexity of the environmental variables that have to be taken into account to understand the distribution and frequency of malaria. The modern experience suggests that the partial drainage achieved temporarily by Claudius would have actually increased the frequency of malaria around the remainder of the lake, especially considering that the climate as a whole was in any case warmer for much of the period of the Roman Empire (see Ch. 4. 5

  below). The theme of unintended side-effects of human activity will recur in the course of this chapter.⁷²

  Mosquito larvae can grow very rapidly, in a few days. They may live in moist ground for three or four days. Once they have pupated, it does not matter if the water of the pool evaporates completely and the ground completely desiccates, since it suits the adult mosquitoes to emerge from completely dry soil or sand, for example sand from the banks of the river Tiber as shown by Celli’s and Grassi’s experiments.⁷³ Gradual desiccation of pools during ⁷¹ Virgil, Aeneid 7.759–60.

  ⁷² North (1896: 117–18) noted the temporary increase in malaria during the modern drainage operations, although by the end of the nineteenth century the Fucine basin had become one of the areas of anophelism without malaria discussed by Hackett and Missiroli (1931). Hare (1884: ii. 190) and Letta (1972: 13 n. 12) also commented on the presence of malaria around the lake in the early modern period.

  ⁷³ Celli (1900: 78); Sambon (1901 a: 199).

  Ecology of malaria

  73

  12. The coastal forest of the Parco Naturale della Maremma, in the direction of the Ombrone river valley, and the Canale Scoglietto Collelungo.

  In the past the water in such drainage canals often flowed too slowly to prevent mosquitoes breeding in them. The pine forest (principally Pinus pinea) on the left was planted in the nineteenth century. The Monti dell’

  Uccellina (to the right and behind the line of sight) are covered by oak forests.

  the summer was assisted by hot, dry south winds, which were associated with malaria by Theophrastus and other ancient authors.⁷⁴

  In the fifth century  Empedocles is said to have blocked up a mountain gorge in order to prevent a pestilential south wind from bringing problems in pregnancy to women (placental malaria) and disease on to the plain surrounding his own city of Akragas in Sicily.⁷⁵ Horace also mentioned the pestilential south wind in his odes.⁷⁶ Similar ideas recurred throughout later history. The doctor Perinto Collodi at Bibbona gave a detailed description of the ⁷⁴ Theophrastus, de ventibus 57, ed. Coutant and Eichenlaub (1975): ka≥ p3lin xhro≥ ka≥ m¶

  Ëdat*deiß Ônteß oÈ nÎtoi puret*deiß.

  ⁷⁵ Plutarch, Moralia 515c: Ø d† fusikÏß ∞Empedokl[ß Ôrouß tin¤ diasf3ga barŸn ka≥

  nos*dh kat¤ t0n ped≤wn tÏn nÎton ƒmpnvousan ƒmfr3xaß, loimÏn πdoxen ƒkkle∏sai t[ß c*raß (It was thought that the natural philospher Empedocles shut pestilence out of his country by blocking a gorge, which allowed an oppressive and unhealthy south wind to blow on to the plains.). See also Plutarch, Moralia 1126b and the other sources cited by Diels-Kranz 31 A1, A2, A14.

  ⁷⁶ Horace, Carmina 23.1–8.

  74

  Ecology of malaria

  association of the sirocco wind with disease among seasonal migrant workers returning from the Tuscan Maremma to Liguria in 1614. Domenico Panarolo (1587–1657) described the austral wind as a ‘deadly enemy of health’. Elsewhere in his works on winds and airs he noted the idea prevalent in Rome at the time that the sirocco wind brought ‘bad air’ to Rome from the Pontine Marshes. Not everyone accepted such ideas. The anonymous author of a tract on mal’aria written in the late eighteenth century perceptively argued that ‘bad air’ was generated regardless of which wind was blowing and that in fact it was most abundant if there was no wind at all (mosquitoes don’t like strong winds). Baccelli maintained that the sirocco wind was unhealthy in Rome in the nineteenth century, particularly if it was humid.⁷⁷ However, a very long period of continuous dry heat during the summer tended to reduce the frequency of cases of malaria, since the mosquitoes eventually began to run out of breeding sites. Consequently the frequency of malaria increased after occasional summer showers, and particularly after the first autumn rains.⁷⁸ This combination of circumstances again illustrates the complexity of the phenomena in question.

  Anopheles larvae generally prefer stagnant water. Lancisi noted that running waters were healthy.⁷⁹ Nevertheless they can also thrive in water that is moving very slowly. A. labranchiae and A.

  sacharovi are happy to breed in ditches or canals so long as the water is not moving faster than about two kilometres per hour. Consequently the construction of canals to drain marshes frequently made the situation with regard to malaria worse rather than better in central Italy in the past. This happened, for example, during the project to reclaim the Tiber delta region around Ostia in 1885–9, since the new drainage channels proved to be even better breeding habitats for Anopheles mosquitoes than the marshes that they drained. Around Ostia Anopheles larvae were found in most of the
drainage canals, which were stagnant and overgrown with aquatic vegetation, by the time of Sambon’s field observations in the ⁷⁷ For south winds as ‘bringers of fever’ (puret*deiß) see also [Aristotle,] Problems 1.23.862a, Pliny, NH 2.48.127, Celsus, de medicina 1.10.4 and 2.1.3–4; Sidonius Apollinaris 1.5.8

  associated the Atabulus wind from Calabria with malaria; Doni (1667: 79–84); Lancisi (1717: 49); Cipolla (1992: 52) on Bibbona; Panarolo (1642 a) and (1642 b): inimico mortale della salubrità; Lapi (1749: 64–5); Anon. (1793: 24); Baccelli (1881: 161–3); North (1896: 138).

  ⁷⁸ Hirsch (1883: 258) noted that an epidemic started in Rome in October 1795 after the first autumn rains, following a long dry summer.

  ⁷⁹ Lancisi (1717: 30–2).

  Ecology of malaria

  75

  summer of 1900.⁸⁰ This is likely to be one reason why the reported drainage of the Pontine Marshes in 160  by Cornelius Cethegus failed to make any impact on malaria and probably even intensified it, since malaria was certainly endemic in this region in the Late Republic, as will be seen in Chapter 6 below. The gradient of the land in the Pontine Marshes was too low for canals to carry water away rapidly. Drainage is a complicated business. Many, perhaps even most, drainage schemes in antiquity were probably failures.

  There are numerous examples of early modern drainage schemes that were spectacular failures with regard to malaria, besides the Ostia project already mentioned. Doni noted that the bonifications (the reclamation by drainage of marshlands) of Pope Sixtus V