Malaria resurgence risk in southern Europe: climate assessment in an historically endemic area of rice fields at the Mediterranean shore of Spain

Change of climate factors and other variables related to environmental modifications included within the broad term of global change have a proven impact on the transmission of infectious diseases [1, 2], caused by different types of infectious organisms including microparasites (viruses, bacteria, rickettsia and protozoans) and also, as very recently proved, metazoan macroparasites (helminths) [3–5], among which mainly vector-borne diseases [6–9], such as malaria [10–13]. Prediction studies indicate that the incidence of infectious diseases is likely to increase as a consequence of climate change and, in the case of malaria, it has been calculated that 50 million new cases will occur by 2100 [14]. Global warming has been considered as a potential risk for malaria resurgence in northern hemisphere areas [12, 15].

Southern Europe is among the most risky regions for malaria resurgence due to (i) the appropriate latitudes with benign climate such as mainly those under mildering influences of the Mediterranean sea, (ii) the closeness to Africa enabling an uncontrolled entry of immigrants from sub-Saharan Africa by means of boats, whether directly through the Mediterranean or indirectly through the Pacific into first the Canary Islands and afterwards the Iberian Peninsula, (iii) the wide presence of anopheline vectors which played an important transmission role during the first half of last century, (iv) the pronounced antropogenic modifications of the environment related to modern development, industry and construction growing giving rise to settlement concentrations consequence of increasing population and tourism which attract a high number of poor immigrants coming from endemic countries and looking for jobs created within the construction and services sectors, and (v) the high immigration rates occurred in recent years mainly from poor endemic areas of developing countries of other continents. All this applies to Spain, where the increase of international travel and immigration from presently malaria-endemic countries has been an exponential phenomenon and global warming has become very evident in the last two decades. In this country, as well as in the neighbouring Portugal, Anopheles atroparvus was the most efficient vector [16, 17]. At present, this anopheline species shows high population densities similar to those observed in old times when there still were authochthonous transmission areas [18, 19]. These facts call for the need to carry out studies on the present potential transmission risk of malaria in areas of Spain, which were historically endemic to assess resurgence probabilities.

In Spain, benign tertian fever caused by Plasmodium vivax and, to a lesser degree, malign tertian fever caused by Plasmodium falciparum and that of quartan periodicity caused by Plasmodium malariae, were endemic until mid last century. In 1943, around 400,000 cases were diagnosed and 1,307 deaths due to malaria were registered. The last autochthonous case was registered in May 1961, and Spain was declared a malaria-free country since it received the official certificate of eradication in 1964 [20]. However, imported cases, mainly stemming from immigrants and tourists, are yearly reported. The surge in travel to exotic countries where malaria is present as well as the increased rate of immigration from endemic countries have led to a steady increase in imported malaria cases from 21 in 1967 to 263 in 1995 and 351 in 2004 [21, 22]. Thus, the possibility that Spanish Anopheles could facilitate the re-introduction of Plasmodium into the country carried by immigrants or travellers who did not follow the appropriate prophylactic measures, has to be taken into account, depending on the receptivity towards its various strains. Currently, new studies are being carried out on the receptivity of the European vector A. atroparvus towards African Plasmodium strains. Previous studies already revealed that it is not susceptible to the afro-tropical P. falciparum strains [6, 18, 23–26], but susceptible to the European strains [6, 27, 28], although probably fully susceptible to infection by P. vivax strains imported from Africa [29]. In addition, Asian or American P. falciparum strains may also be imported [29]. Hence, a transmission resurgence risk, as reflected by the detection of possible autochthonous cases in south European countries as Spain [29], Greece [30, 31], Italy [32–34] and France [35, 36], and even more northward as in Germany [37], should be taken into account. These facts make it necessary to carry out studies to analyse the present situation and to assess whether the ecological, entomological and social characteristics of the historically endemic areas in Spain allow for the transmission of the disease nowadays. Studies with similar objectives have very recently been carried out in other European countries such as Italy [33, 38], France [39–43], Portugal [17] and Germany [44].

A field analysis of potential risk factors of malaria transmission resurgence has so far never been made in Spain. A historically endemic area as the Ebro Delta, at the Mediterranean shore (Figure 1), appears to be the best way for such a study. Malaria in the Ebro Delta was first reported in Tortosa in 1738 and between 1800 and 1900 the disease was included among the priority treatment campaigns prescribed by local physicians [45]. The problem of malaria in the area began to worsen from 1860 onwards after the introduction of rice cultivation [46]. The rapid expansion of this crop together with the bad planning of drainage systems, the precarious conditions of day labourers and deficient hygienic conditions caused an increase of the disease and turned it into a serious health and socioeconomic problem, as a consequence of the decrease of day labourers and the costs incurred by treatments and anti-malaria campaigns. Old data showed that 26,557 out of 70,540 inhabitants of the Ebro Delta were exposed to a paludic environment and that the number of infected people rose from 6,207 in 1912 to 12,545 in 1915, reaching an outstanding 95% in some municipalities of the area [47]. Plasmodium falciparum and P. vivax were the causes of two thirds of all cases [45]. There has been no indication of malaria transmission since its eradication. Thus, the Ebro Delta presents nowadays a situation of anophelism without malaria, with A. atroparvus as the only anopheline species present [16]. This feature shall be highlighted when compared to old malaria endemic areas and presently potential resurgence areas in other Mediterranean countries as France and Italy, where many different Anopheles species coexist in the same area and a different vector species dominate, as Anopheles hyrcanus in France [39, 40] and Anopheles labranchiae in Italy [33, 34, 38]. Anopheles atroparvus is an halophilous species following seasonal population dynamics nearly throughout Europe. Changes in climatic conditions, including temperature, evapotranspiration and surface runoff, all key factors to determining mosquito abundance and survivorship [48], may increase the favourable period for the development of the vector and the transmission of the disease.

With the aim to assess whether the present ecological characteristics of the Ebro Delta permit malaria transmission nowadays, the A. atroparvus population has been monitored and climatic as well as terrain conditions have been characterized to determine if this parasite/vector association could satisfactorily complete its biological cycle. This assessment has been carried out by means of a climatic analysis and applying the Gradient Model Risk Index based on climatic conditions favouring the development of the parasite and vector [49, 50]. For terrain analysis, satellite imagery, known to furnish valuable data for the forecast of malaria [51], has been used by means of remote sensing information selected to characterize the overall land cover of rice fields of the Ebro Delta as data from the Eurasian Land Cover database [52], and the Normalized Difference Vegetation Index (NDVI) which has been proved to be highly efficient for prediction studies in water-related parasitic diseases [53].

The analysis of the various classifications of the Land Cover database shows that in the Ebro Delta irrigation farming, namely rice, predominates, which is shown at the end of April and the beginning of May and reaches its maximum growth in July and August, respectively. Monthly NDVI values reach their maximums in July and August coinciding with the dry period in the Ebro Delta, reflecting the growth of rice characterized by irrigation through artificial flooding. Thus, the rice fields in the Delta provide an ideal environment for the development and breeding of A. atroparvus owing to the existence of great masses of clean, motionless water, as already noted long ago [62, 63]. Rice cultivation periods, moreover, coincide with temperatures permitting the completion of the vector cycle, as shown by the detection of A. atroparvus during all the months studied. Mosquito maximum abundance in June, July and August also coincides with NDVI maximum values as well as with maximum growth in the rice fields. The importance of cultures of this kind of crop in the development of malaria vectors has already been noted in other south European countries [33, 38, 42, 64, 65].

A potential malaria transmission risk in the Ebro Delta is revealed when considering the climate data and the characterization of the study area along the 26-year period analysed. The climate diagram indicates that favourable temperature conditions can be found between May and September for A. atroparvus and P. falciparum, and such a favourable period extends from April until October for P. vivax. The results of the GMR index place the transmission risk solely in June, September and October for P. vivax and in September for P. falciparum. The calculation of this index makes it possible to define the period of potential transmission risk, considering climatic factors only, but in a much more precise manner, taking also into account the minimum temperature required for parasite and vector development as well as the presence of rain water needed for the completion of the A. atroparvus life cycle. Nonetheless, the insufficiency of water needed for vector development in the Ebro Delta, typical for the dry period between June and August, is compensated by the artificial flooding of the rice fields. Therefore, the transmission risk period should be extended to May and September for P. falciparum and to May and October for P. vivax.

Relative humidity is another essential factor in malaria transmission. Although Plasmodium parasites are not affected by relative humidity, anopheline mosquito activity and survival are. If the average monthly relative humidity is below 60%, it is believed that the lifespan of the mosquito is shortened to such an extent that malaria transmission does not occur [66]. The climatic analysis of the study area shows that relative humidity is above 60% along the entire year, bar April. Thus, suitable conditions for the development of A. atroparvus and Plasmodium turn the Ebro Delta into an ecosystem with favourable climatic characteristics for malaria transmission.

The results obtained agree with old observations [45], which noted that 99% of all new malaria cases in the Delta were detected at the end of summer, that the first cases tended to appear in July, and that epidemics occurred in August, particularly during the last three weeks of that month with the detection of numerous cases. Our results reveal that the favourable transmission period is currently longer than it was, starting two months before, in May, and lasting until September in the case of P. falciparum and until October in case of P. vivax, respectively. The discovery of larvae in a different place from a rice field (the typical breeding site of A. atroparvus in the Ebro Delta) and the discovery of males in May (up until then no larval stages of Anophelinae and Culicinae had been detected in the rice-fields of the Delta) support the prolongation of the malaria transmission risk until October, although the rice cultivation period is completed by then.

Climate data for 2005 and 2006 show that although an increase of temperatures took place, favouring an extension of the transmission period, at the same time a decrease in precipitation occurs, resulting in an extension of the dry period, which in turn would a priori negatively affect the favourable period for vector development if artificial irrigation water would not be available. Population dynamics of A. atroparvus populations in the Ebro Delta follow a typical thermophilic trend where rice fields ensure a constant presence of water. The highest values were recorded in the period from June until August, when the abundance of the population peaked. Lowest values were recorded in September, but a new peak appeared in October-November. This seasonal dynamics is quite similar to those presented by A. labranchiae in Italy [34] and A. hyrcanus in France [40], both assessed during the same biannual period (2005-2006).

Transmission risk analysis for 2005 indicates that conditions for vector development of the two Plasmodium species considered are still suitable in the May-September period for P. falciparum and in May-October for P. vivax. Although the GMR index shows, in the case of P. vivax, a disappearance of the transmission risk in June, it has to be kept in mind that the rice field irrigation counteracts the lack of rain water in that month and consequently the transmission risk persists.

Climate data analysis for 2006 confirms that the increase in temperatures prolongs the favourable period for both vector and P. vivax development, lasting from May until November. Although the climate diagram shows that the risk period starts in April (favourable temperatures), results derived from the GMR index and the fact that rice is not grown in that month, suitable conditions for the completion of the Anopheles life cycle do not exist, and therefore April can not be considered within the potential risk period. However, the GMR index did not indicate a P. falciparum transmission risk in 2006 as a consequence of reduced precipitation in that year. Thus, P. falciparum sees its favourable period limited to May until September, when the rice field irrigation counteracts the lack of rain water required for the completion of the Anopheles life cycle, and temperatures, following the climate diagram, facilitate parasite and vector development.

The evolution of the GMR index shows that the increase in temperatures does not actually mean an increase in the malaria transmission risk if this is accompanied by a decrease in precipitations. Although temperatures favour parasite development, the lack of water prevents vector development. This fact reduces the number of parasite generations as well as the transmission period. Nevertheless, this limitation is offset by the artificial flooding of the rice fields in the Delta. Therefore, the use of this index, which solely considers the existence of breeding sites and periods as a consequence of precipitation, has to be combined with a spatial characterization of the study area, which helps to detect breeding sites and periods. NDVI from remote sensing sources proves to be a good marker for such a purpose when dealing with a rice field area.

The results obtained make it possible to conclude that the Ebro Delta currently presents ecologically favourable characteristics for the re-appearance of malaria if an appropriate malaria strain would be introduced and that the month window period for potential transmission has even widened due to present global warming impact when compared to last century when this area was endemic. This fact makes it necessary to verify whether the Plasmodium strains potentially introducible by infected immigrants, tourists or travellers arriving to Spain could be compatible with the European A. atroparvus. Additionally, factors that might influence vector population dynamics as well as the transmission of Plasmodium (blood preferences of the vector, aggressiveness, gonotrophic cycle, etc.) have, therefore, to be analysed in the Ebro Delta area in question.