Malaria and Rome: A History of Malaria in Ancient Italy

by Robert Sallares

  Theophrastus’ description of the region around ancient Circeii shows that he was aware of ongoing environmental change in the form of the alluviation which was thought to have attached Circeii, regarded in antiquity as once having been an island, to the mainland of Italy. Monte Circeo is regarded by modern geologists as e”nai ]∫Îna. t[ß d† n&sou tÏ mvgeqoß per≥ øgdo&konta stad≤ouß . . . t0n g¤r ƒn t∫ Lat≤n7

  kal0n ginomvnwn Ëperbol∫ ka≥ t0n ƒlat≤nwn ka≥ t0n peuk≤nwn—me≤zw g¤r taıta ka≥

  kall≤w t0n ∞Italik0n. (The whole territory of the Latins is well-watered; the plains contain forests of bay, myrtle, and wonderful beech. They fell timbers of it so long that they span the entire length of the keel of an Etruscan ship. The hills have forests of fir and silver-fir . . . The so-called Circaion is an elevated promontory, but it is densely wooded and has oak trees, a lot of bay, and myrtle. The land surrounding the Circaion has been created recently by sedimentation from certain rivers, but the Circaion was formerly an island. The island was about eighty stades in circumference . . . silver-fir and fir grow extremely tall in Latium and are taller and finer than in southern Italy.).

  ¹⁶⁰ Quilici (1979: 76–87) on the ancient forests of the Campagna Romana, cf. Pratesi and Tassi (1977: 76); Traina (1990: 16); Grandazzi (1997: 65–73) on the environment of Latium.

  Meiggs (1982: 219, 243–5) and Fraser (1994: 184–6) discussed Theophrastus on the forests of Latium without considering the problem of the beeches. The inaccuracy of Theophrastus’

  information (if it is not simply the case that the text has become garbled during manuscript transmission) fits Fraser’s emphasis on the paucity of information available to Theophrastus from the western Mediterranean in contrast to the large volume of data yielded by Alexander’s expedition into the Persian Empire. The new Budé edition by S. Amigues, with reference to Theophrastus’ description of the beech tree in HP 3.10.1 (cf. 3.11.5), noted that ‘ la distinction entre hêtre de montagne et hêtre de plaine n’a pas de valeur scientifique’. Amigues discussed the ecological adaptability of the beech tree, but did not consider the possibility of climatic change affecting its distribution. Pratesi and Tassi (1977: 38) described ‘ il faggio’ as ‘ un albero che, nel Lazio, vive sulle montagne appenniniche a un’altezza variabile tra i 900 e i 1800 metri’, although they noted that it sometimes descends to 400 or 500 metres above sea level in favourable locations, cf. Leonardi and Menozzi (1995); Blasi et al. (1999). Bietti Sestieri (1980: 9) suggested that the lower limit of the beech tree has been driven upwards by intensive agriculture in the lowlands. Rendeli (1993: 132) noted finds of beech wood in southern Etruria in Bronze Age levels at Sorgenti della Nova, Grotta Misa, and Lago di Mezzano, and in Iron Age levels at Castro and Acquarossa. These sites are all situated close to suitable uplands for beech forests. Magri (1999: 193) noted that there is fossil evidence for Fagus in the lowlands in earlier geological periods at Torre in Pietra, west of Rome. Theophrastus, HP 3.17.1 discussed another tree, Quercus suber, in Tuscany.

  Ecology of malaria

  107

  18. The northern slopes of Monte Circeo, with the forest of the Quarto Freddo, viewed from near Torre Paola.

  simply an isolated part of the Apennines. It is not now thought to have ever been an island during the Holocene period, but there is no doubt that there was substantial alluviation in the adjoining Pontine plain in classical antiquity.¹⁶¹ The promontory of Circeii had dense forests of oak, bay, and myrtle. The remnants of the Pontine forest are still visible on the northern slopes of Mt. Circeo itself and in the adjacent Selva del Circeo in the Parco Nazionale del Circeo.¹⁶² Strabo (quoted in Ch. 6 below) shows that the area of the Pontine Marshes in the vicinity of Monte Circeo was severely affected by malaria, although Monte Circeo itself was above the range of malaria in the past, since it reaches an altitude of 541

  metres above sea level. Doni regarded Mt. Circeo as healthy in the ¹⁶¹ The Dutch archaeologists from Groningen identified one specific episode of alluviation (Attema et al. (1999) ). They found evidence for a mass mud flow from the valley of Vado La Mola into the Pontine plain, near the important proto-historic settlement of Caracupa Valvisciolo near Sermoneta. They attributed this mud flow to ‘vegetational denudation of hillslopes’.

  ¹⁶² Ovid, Metamorphoses 15.716–17 described Circeii as surrounded by marshes. Pliny, NH

  3.5.57–8 simply followed Theophrastus’ account of Circeii, cf. Procopius, BG 1.11.2–4. Pliny, NH 19.40.134 said that the beets of Circeii were particularly noteworthy. [Aristotle,] peri thaumasion akousmaton 78.835b–6a believed that a deadly plant-derived drug could be obtained at Circeii. Pratesi and Tassi (1977: 145–8) and Stanisci et al. (1998) described the forests of Circeii as they are today.

  108

  Ecology of malaria

  seventeenth century, while Martial described Terracina as healthy in antiquity. Demetrius used to hunt wild boar in the forests of Circeii.¹⁶³

  By the Late Republic the expansion of the city of Rome required very large quantities of timber for construction and other purposes.

  Strabo states that the wood for buildings in Rome came mainly from Etruria. It was transported along the Tiber and its numerous tributaries to the city. Pliny noted that the Tiber had forty-two tributaries below its junction with the Chiana. Strabo also mentions the region of Pisa as providing timber for buildings at Rome and for shipbuilding. Rusellae, near Grosseto, was one of the towns which offered timber for shipbuilding to Scipio Africanus in 205  for his expedition to Africa. Another cause of deforestation was the requirement for firewood for smelting metals (e.g. around Populonia and in the Colline Metallifere).¹⁶⁴ The breaking up of previously closed forest environments in the lowlands favoured Anopheles mosquitoes, some species of which prefer pools of water that are exposed to sunlight for breeding purposes. In the early eighteenth century  Lancisi observed that stagnant waters, which faced the open sky in summer, were particularly dangerous; such waters had lots of insects and abundant vegetation (which provides cover for mosquito larvae). A. sacharovi particularly likes inland waters which are open to sunlight in summer. Lancisi attempted to prove his observations by practical experiments:

  It is impossible to ignore the fact that in summer rainwater kept in an open vase becomes polluted with mosquitoes far more quickly and abundantly than anything else.¹⁶⁵

  ¹⁶³ Doni (1667: 104); Martial 5.1; Polybius 31.14.2–3, ed. Buettner-Wobst.

  ¹⁶⁴ Strabo 5.2.5.222–3C; Pliny, NH 3.5.53–4; A. Betocchi, Del fiume Tevere, in Monografia (1881: 206–7) listed the tributaries of the Tiber; Livy 28.45.18 on Rusellae; Sallares (1995) and various papers in Frenzel (1994) considered deforestation in antiquity. The demands of ‘industry’ for timber in the past should not be underestimated. De Felice (1965: 86) gives interesting statistics, which can be scaled upwards for antiquity, about the consumption of charcoal by craft industries in the city of Rome in the eighteenth century, when the city’s population was much smaller than during the Roman Empire: ‘ la sola città di Roma consumava annualmente 12 milioni circa di kg di carbone’. Meiggs (1982: 218–59) discussed the timber requirements of the city of Rome in antiquity and emphasized the need for charcoal for heating, industry, cremations, and hot water in the public baths, cf. Rausing (1987). Toynbee (1965: ii. 585–99) is still interesting on the question of deforestation.

  ¹⁶⁵ Lancisi (1717: 58–9, cf. 26–9): neque vero hic dissimulandum est, aquam pluviam per aestatem, si aperto in vase detineatur, culicibus longe citius, et copiosus quam quaevis alia foedari. Tommasi-Crudeli (1881 b) wrote a short article about plant pots inside the house as a source of ‘bad air’; Russell (1943: 60).

  Ecology of malaria

  109

  The first millennium  witnessed a substantial shift away from the closed forests of the Neolithic period towards the modern open countryside. Dionysius of Halicarnassus described one famous closed forest, the Sila in Bruttium, which he reg
arded as a source of timber for the whole of Italy.¹⁶⁶ Such forests were under attack in Roman times. Deforestation in upland areas of central Italy led to increased run-off of rainwater. Pliny the Younger noted that at Tifernum in Umbria perennial streams watered beautiful meadows full of flowers, but the ground, because it was sloping, was not marshy, as water drained off into the Tiber. The water ended up elsewhere. A long series of major floods occurred in the city of Rome in antiquity, recalling the catastrophic flood of November 1966 at Florence, Grosseto, and elsewhere in Tuscany, which occurred after forty consecutive days of heavy rainfall. The construction of the Cloaca Maxima suggests that a need for drainage was already felt from the very beginning of Roman history. Strata dating to the imperial period in the Roman Forum are now covered by six or seven metres of alluvial deposits and are themselves several metres above the levels of the archaic period.¹⁶⁷ Pliny the Elder states that the Tiber floods were nowhere worse than in the city of Rome itself, and Pliny the Younger gives a graphic description of the effects of a Tiber flood. Orosius mentions a devastating flood in 241 . Cassius Dio recorded that the lower parts of Rome were completely flooded by the Tiber in 54 , causing severe damage to buildings. He attributed the flood to very high rainfall upstream, or to the sea driving back the river water. Both explanations are possible. Another Tiber flood made the city of Rome navigable by boat for three days in 23 . The following year, Rome was yet again submerged. Tacitus mentioned a severe flood in  15, which again caused severe damage to buildings and loss of life, and described an interesting debate in the Senate about these problems. In the end the Senate, whether because of the difficulty of water-management enterprises, or the inevitability of protests from communities liable to be affected by river diversions, or because of superstitious beliefs, decided to do nothing. This illustrates the large degree of helplessness of the Romans in the face of ¹⁶⁶ Dionysius Hal., AR 20.15, ed. C. Jacoby (1891): lh skierÏn åpoteloısa di’ Òlhß Ómvraß tÏ Ôroß (the forest keeps the mountain shaded throughout the day). Douglas (1955: 228–30) described the Sila forest as it is now, and Béal (1995) reviewed all the evidence from antiquity.

  ¹⁶⁷ Quilici (1979: 69–70).

  110

  Ecology of malaria

  these ‘natural’ (albeit unintentionally partly man-made) phenomena. The flood in  15, the year after the death of the emperor Augustus, shows that the measures which he is said to have taken to control the Tiber were completely ineffectual. Tacitus records another severe flood in  69. Terrible floods also occurred in Rome in the reigns of Nerva, Trajan, and Marcus Aurelius. Cassius Dio mentions another flood in  217, and there were doubtless others which have gone unrecorded.¹⁶⁸

  The records of the worst floods in Rome during the last millennium show that they principally occurred during the months of September to February. As Rome gradually dried out after these great floods, mosquitoes would have found countless ideal breeding sites within the perimeter of the city of Rome itself. Literature from the early modern period shows that Tiber floods were regarded as a very important factor in the generation of malaria in the city of Rome. The Tiber was not enclosed by the Lungotevere river walls then as it is today. A huge flood in late December 1870, a few weeks after Rome had become the capital of the newly unified Italian state, put the question of taming the river Tiber at the top of the political agenda for the new government. The districts of the city worst affected by malaria were along the river. Floods in September and October increased mosquito populations and the intensity of malaria immediately. Floods later in winter would drench the ground and make it more suitable for mosquito breeding purposes in the following summer. A description of a medieval flood in December  791 makes it clear that much water remained in the city long after the Tiber had ceased to flow through it. It was not coincidence that pestilence accompanied the floods of 23–22

  , according to Cassius Dio, although many other diseases besides malaria would have been active as well, especially waterborne intestinal diseases, which can interact with malaria in ways which will be described later (see Ch. 5. 2 below).¹⁶⁹

  ¹⁶⁸ Le Gall (1953: 29–34) discussed the evidence for Tiber floods in antiquity. The following sources are noteworthy: Pliny, Ep. 5.6.11–12; Pliny, NH 3.5.55; Pliny, Ep. 8.17; Orosius, Hist. 4.11.5–7, ed. Arnaud-Lindet (1991); Cassius Dio 39.61.1–2, 53.33.5, 54.1.1, also 56.4 for a flooding of the Campus Martius by the Tiber in  12, and 79.25.5; Suetonius, Augustus 30; Tacitus Annals 1.76, 79 (with LeGall (1953: 120–5) ); Tacitus Histories 1.86; epitome de Caesaribus 13.12–13, ed. Pichlmayr (1911); scriptores historiae Augustae Hadrian 21.6, Antoninus Pius 9.3 and Marcus Aurelius 8.4; Claudian, de bello Gildonico 39–43; Quilici (1979: 66–8).

  ¹⁶⁹ Bocquet (1998) described the events of 1870. A. Betocchi Del fiume Tevere in Monografia (1881: 244–5) listed the heights and months of the worst Tiber floods in the last eight hundred years. Baccelli (1881: 156, 159–60) regarded Tiber floods as very important in generating Ecology of malaria

  111

  Besides floods, heavy rainfall in winter and spring was also beneficial for mosquito breeding. A correlation between heavy rainfall in winter or spring and an elevated frequency of malaria was observed in Rome in the last century. North described as an example the events of 1879, when exceptionally high rainfall in the period from March to May was followed by a severe epidemic of malaria in the summer and autumn of that year. Similarly Bellincioni observed a very close correlation at Grosseto between rainfall levels in the months from September to May and the frequency of malaria in the following summer and autumn over a period of thirty-two years that was studied, from 1900 to 1932. He explained the connection in terms of fluctuations in the level of the water table, which must be linked with fluctuations in the frequency of malaria via fluctuations in the size of mosquito populations, which in turn depend on the size of suitable habitats available to them for breeding purposes.¹⁷⁰ Pliny the Younger observed that, wherever one dug, water was found at his villa near Laurentum, indicating a very high water table.¹⁷¹ Many of the coastal areas, which were the most severely infested with malaria in the medieval and early modern periods, were expanding in size in Roman times because of alluviation and were prone to flooding. For example, the Etruscan cities of Vetulonia and Rusellae, near the site of modern Grosseto, were much closer to the sea in antiquity than they are today.¹⁷² They were situated by a gulf which in Roman times became a brackish lagoon, the lacus (or amnis) Prilius. Today the area of the former lake forms part of the plain of Grosseto, but it existed into the early modern period as the Lago di Castiglione, a freshwater lake sometimes extending over more than a hundred square kilometres. The history of this area is a classic example of the spread of malaria around the margins of a coastal lake which malaria in Rome. He reckoned that malaria occurred where the river could deposit sediments on its banks, but not where there were river walls. Davis (1992: 171) and (1995: 84, 179–81, 209–10) recorded the Tiber floods in December 791, November 844, January 856, and October and December 860 .

  ¹⁷⁰ North (1896: 145–7); Bellincioni (1934). Hay et al. (2000) observed a three-year cycle of P. falciparum epidemics which could not be correlated with climatic trends in a highland region of western Kenya, and suggested that intrinsic population dynamics offer the most parsimonious explanation of the intervals between epidemics. However the evidence from Italy strongly supports climatic explanations for malaria epidemiology in Europe in the past.

  ¹⁷¹ Pliny, Ep. 2.17.25.

  ¹⁷² Dennis (1878: 222–34) on Rusellae, which is situated about 170 metres above sea level, while Vetulonia is around 300 metres above sea level.

  112

  Ecology of malaria

  developed and then gradually became cut off from the sea.¹⁷³ The flood plain of the river Ombrone was too flat for drainage purposes, like the Pontine plain. In the early modern period the increasing use of quinine play
ed a more important role in reducing malaria in the province of Grosseto than the bonifications.¹⁷⁴ However, the final blow to malaria in Italy as a whole was delivered by the insecticides DDT (against adult mosquitoes) and Paris Green (against mosquito larvae) during and after the Second World War, for example during the operations to defeat the terrible epidemic in the Monte Cassino region. Consequently the balance of probability is that drainage operations in Roman times, when quinine and DDT were not available, had limited effects on malaria, just like their early modern counterparts. Anopheles mosquitoes still exist in the Maremma in spite of the bonifications.¹⁷⁵

  Besides deforestation in the uplands, intensive farming in the vicinity of Rome to feed the burgeoning urban population also played a role in soil erosion, which increased enormously in Roman times. The rate of soil erosion in Latium increased ten times in the second century  compared to its previous rate. This may be attributed to dense settlement around the city of Rome, as suggested by the Capena and Veii archaeological field surveys, and intensive market gardening to produce food for the increasing urban population.¹⁷⁶ We shall see in Chapter 8 below that there was a strong causal connection between gardening and malaria.

  ¹⁷³ Pinto (1982: 13–14).

  ¹⁷⁴ Ciuffoletti and Guerrini (1989: 67, 95, 136).

  ¹⁷⁵ Raffaele and Coluzzi (1949); Garaci (1947); Merzagora, Corbellini, and Colluzi (1996).

  Alberto Coluzzi led the operations against this epidemic. Malaria epidemics are attested in the Monte Cassino region and the valley of the River Liri towards Frosinone as far back as the early medieval period: e.g. chronica monasterii Casinensis 1.22, ed. H. Hoffmann (1980), Die Chronik von Montecassino ( Monumenta Germaniae Historica. Scriptores, xxxiv): Qui videlicet sanctus vir cum depositus fuisset septimo idus Octobris, in loco, quo reconditus est, multos febre detentos diversisque lan-guoribus oppressos, ex fide poscentes ad suos cineres precibus ac meritis suis pristine saluti restituit (This undoubtedly holy man [sc. abbot Deusdedit], when he was buried on 9 October [834 ], in the place, in which he rests, restored to their previous good health many people gripped by fevers and oppressed by various kinds of weariness, who in accordance with their faith invoked his ashes with prayers and their own services.); 2.96 Sed cum predictus apostolicus Romana febre iam dudum langueret, circa ipsam natalis Domini festivitatem adeo graviter infirmatus est, ut pro certo se mori putaret [December  1057] (However since the above-mentioned bishop had already long been severely afflicted by Roman fever, he was so ill about the time of the festival of the birth of Christ that he thought that it was inevitable that he was going to die.), cf.