In recent years, several vector-borne diseases have re-emerged and spread in Europe with major health, ecological, socio-economical and political consequences. Most of these outbreaks are linked to global and local changes resulting from climate change, human induced landscape changes or human population activities. Malaria, a former European endemic disease recently struck European boundaries in countries such as Azerbaijan, Georgia and Turkey where it had been eradicated after World War II [1]. Only a few autochthonous cases have been reported recently in Europe, but these events indicate that the malaria situation needs to be re-examined in Europe [1–5].

In metropolitan France, malaria was endemic until the beginning of the 20^th century in marshy areas such as the Landes, the Dombes, Brittany, Alsace, the Rhone delta, Roussillon and Corsica [6]. Then, it decreased drastically due to the drying of marshes, growth of livestock, improvement of housing and life conditions and the use of quinine. The last outbreak was observed in Corsica from 1966 to 1972, with about 30 Plasmodium vivax cases [7, 8]. Malaria disappeared from the Camargue after World War II [9].

Nowadays, all the malaria cases reported in France are only imported cases [10], excepted for three suspected, but not-confirmed, autochthonous cases in 2006 [11](Doudier, unpublished data). In 2004, the total number of imported cases was 6,109 with a predominance of Plasmodium falciparum [12]. Between 1977 and 2000, 28 airport malaria cases following infected mosquito importation were recorded [13] and very few cases have been notified as congenital malaria or accidental blood exposure [10].

Thirteen Anopheles species have been reported in metropolitan France [14]. Among them, two species were considered to be primary vectors because of their abundance and their potential anthropophily: Anopheles atroparvus in continental France and Anopheles labranchiae in Corsica.

Despite drying of some marshes and consecutive reducing of mosquito populations during the 20^th century, Anopheles mosquitoes are still present in France and could be very abundant in some places generating an "anophelism without malaria" situation. However, global and local changes may modify Anopheles biological parameters linked to malaria transmission (vector density, contact between humans and vectors, longevity, species). Only few data on anopheline potential vectors have been collected in France since the seventies [15].

An in-depth longitudinal survey was conducted in the Camargue, to assess vector species and distribution and evaluate mosquito vectorial capacity related to human malaria transmission risk. Dynamics, feeding preferences, parity rates and nightly activity were studied for different Anopheles species.

Anopheles hyrcanus was the most collected species in both areas and particularly at Marais du Vigueirat where thousands of specimens were captured per night during the peak. Rice fields constitute prolific larval sites for this species and large populations are frequently associated with some rice growing areas as described in Turkey or Greece [20]. Although it was much less abundant than An. hyrcanus, An. melanoon was predominant within the Maculipennis complex in the two study areas. Rioux noticed this abundance by the years 1950: An. melanoon was very abundant in the coastal area and paddies constituted its favourite breeding site [21]. Salières confirmed these observations 20 years later [15]. Data presented here show that An. melanoon abundance has not diminished in the Camargue from the years 1950. Anopheles atroparvus was very rare in 2005, although it had been described as the major vector in the past. It was very abundant and present in several environmental patterns of South East France by the 1940's and 1950's [9, 21], whereas Salières captured it only five times during a 4-year survey 20 years later [15]. Observations reported here confirmed that this species remained very rare.

Breeding sites of An. algeriensis are constituted by resurgence of water tables [21]. This kind of biotope is located on the east side of the Marais du Vigueirat where the "Crau" water table appears. This is why this species was captured only at Marais du Vigueirat and particularly in spring. The absence of An. algeriensis in summer could be explained by resurgences being dry during summer due to drought in the South of France.

Anopheles maculipennis s.s. was reported several times in France, but it is associated with fresh water [22]. It was reported only twice near the south coast of France and in fresh water biotopes [15, 21]. Moreover, 16 An. maculipennis complex specimens were collected in a horse bait trap in 2004 at Lunel-Viel [23]. Six out of 11 specimens processed by PCR were identified as An. maculipennis s.s. Lunel-Viel is located at the outer limits of the Camargue, in a dry area without any marshes, paddies or pools submitted to salt water table resurgence. The absence of An. maculipennis s.s. in results presented here is in accordance to these environmental differences.

Anopheles claviger s.s., Anopheles petragnani and Anopheles plumbeus have been reported several times in south-east France. They were not captured during the survey conducted in 2005 because their ecological niches were rare or absent of the two sites and this study did not focus on their breeding sites (which correspond to tree holes for An. plumbeus, rivers stream and small, cold and fresh water collection for the others) [14, 21, 24, 25]. However, An. plumbeus and An. claviger, which were considered as secondary vectors, could be more abundant in some others places, and particularly places close to humans.

Anopheles labranchiae has not been recorded in the Camargue and continental France so far, while it is abundant in Corsica, where it is reported in breeding sites such as small pools with fresh water and marshes [14, 26, 27]. It is also abundant in Italy, particularly in paddies and in rivers and streams [28]. In the context of global warming, distribution of this species has not yet expanded to the Camargue despite abundance of potential breeding sites, such as paddies.

Anopheles hyrcanus and An. melanoon presented similar dynamics in the two areas in 2005, although total mosquito numbers were very different. Their populations began increasing in the middle of June, reaching a peak near the middle of August and decreasing drastically in the middle of September (although An. melanoon decreased earlier). Dynamics of these species collapsed brutally at the end of July in the two areas without any identified cause (such as wind, temperature, treatment by mosquito pesticide, hygrometry, water supplying of breeding sites, significant modification of Anopheles breeding sites or moon cycle).

Populations of resting An. atroparvus remained very low until September or October and then presented a peak in October and November although the number of collected specimens remained small. During this period, more An. atroparvus were captured at Carbonnière and particularly in one horse shelter. In 1943, Sautet had observed that this species was very abundant in September suggesting a near disappearance of this species since the 1950's [9].

Anopheles hyrcanus presented a huge anthropophily in both areas with spectacular aggressiveness on humans: during this study scientists underwent massive attacks from females of this species. Results from others authors confirm the high level of anthropophily in France in the past or in South Eastern Europe and South Western Asia [20, 21, 27, 29]. It was reported as one of the principal mosquito pests in Northern Greece in 2001 [29]. Unfortunately, only nine An. hyrcanus bloodmeals were processed because very few blood-fed An. hyrcanus females were captured in light traps. None of these nine blood meals were human demonstrating an opportunistic trophic behaviour of this species.

In contrast to An. hyrcanus, An. melanoon only exceptionally bite humans. This is confirmed by blood meal analyses: An. melanoon bite animals, especially big mammals (horses and cows). The high degree of zoophily had already been reported [15, 22, 27, 30].

Parity rates were difficult to assess because sampling methods did not always provide enough mosquitoes suitable for dissection (too few or too dry mosquitoes). When it was possible to determine parity rates based on a robust number of specimens and using the same sampling method (light traps), parity rates were reported as inversely connected to population dynamics. This situation was observed at Marais du Vigueirat for An. hyrcanus (June, July and September), An. melanoon (July and September) and at Carbonnière for An. hyrcanus (July and September). Low parity rates, related to young populations, reflect the growth of populations during summer to reach a peak in August, while higher parity rates in September were related to older and decreasing populations.

In the Camargue, a former malaria-endemic area, An. hyrcanus seems to be the only Culicidae likely to play a role in malaria transmission in view of its abundance and anthropophily. However, the experimental susceptibility of An. hyrcanus to infection with P. vivax and P. falciparum strains should be tested for assessing risks for potential transmission of this anthropophilic species. Finally, the «anophelism without malaria» situation is still going on in metropolitan France since malaria elimination, although there is a possibility for local transmission.