Fig 1.
Map of the study area with an ortophotograph of the area.
Ecological units are described in Material and Methods. Urban areas correspond to the villages situated at the border of the study area. White circles indicate the locations of the traps.
Fig 2.
Monthly number of captures of the different mosquito species from March to November 2010 in the different ecological units studied.
Cx. theileri was more abundant in marshland, but also in other landscape units. Cx. perexiguus was more abundant in ricefields, scrubland and other landscapes. Cx. modestus was more abundant in marshland and in June. Cx.pipiens was abundant in marshland, sand dunes and scrubland, Oc. caspius was abundant in all landscapes, while Oc. detritus was significantly more abundant in halophytic marshes of sand dunes and fish ponds.
Fig 3.
Spatial interpolation by kriging of annual mosquito abundances of: total female mosquitoes (a), Cx. theileri (b), Cx. pipiens (c), Cx. perexiguus (d), Cx. modestus (e), Oc. caspius (f), Oc. detritus (g), and An. atroparvus (h) with an ortophotograph of the area.
Table 1.
Number of female mosquitoes captured in each ecological unit.
Fig 4.
Results of the GLM binomial analysis of the influence of Hydroperiod and NDVI on annual mosquito presences in the five different buffers.
Table 2.
Results of the Generalized Linear Model (binomial) of annual presence for six mosquito species (Cx. theileri is not analysed) vs hydroperiod and NDVI estimated at five buffers of different size.
NDVI2 is the quadratic effect of NDVI.
Fig 5.
Relationships among annual hydroperiod, NDVI and Shannon diversity index and mean annual mosquito species richness.
Fig 6.
Relationships among annual hydroperiod, NDVI, Cx.theileri, Cx. pipiens, Cx.modestus, Cx.perexiguus and Oc.caspius annual abundances.
Table 3.
Results of the Generalized Linear Model (normal) of diversity and mosquito species richness in relation to Hydroperiod and NDVI estimated at five buffers of different size; and results of the Generalized Linear Model (negative binomial) of annual female abundance for five mosquito species in relation to Hydroperiod and NDVI.
Fig 7.
Relationships between Shannon diversity Index and monthly NDVI, and between monthly mosquito abundance and inundation area.
Table 4.
Results of the Generalized Linear Model (binomial) of monthly mosquito presence for seven mosquito species vs Inundation surface and NDVI estimated at five buffers of different size.