3. RESULTS

3.1 DDT resistance status

Resistance to DDT 4% observed in all studied sites showed mortality rates ranged from 10% to 80% during 2008 and 2009 (Fig. 2). Koubri, a conventional cotton cultivation area close to the capital Ouagadougou remained an area where a high rate of mortality was detected (80%) (Fig. 2a).

3.2 Permethrin resistance status

The data recorded with permethrin in 2008 showed resistance (mortality rates less than 90%) in the most of sites investigated whatever the pest control strategy. During 2009, overall, \mortality rates were less than 90% in all study sites, confirming resistance to permethrin (Fig. 3b).

3.3 Deltamethrin resistance status

Whatever the cotton pest control strategies adopted from 2008 to 2014, mortality rates observed with deltamethrin 0.05 % in most An. gambiae s.l. populations were below the 90% threshold indicating high resistance status of mosquito populations. Apart from that, two sites in 2013 have a suspected resistance status with a mortality rate higher than 90% but lower than 98%: Soumousso (95%), Orodara (94%) from Conventional cotton growing area (Fig. c). Only, in 2008 and 2013, Koubri from Conventional cotton growing area, exhibited a susceptibility to deltamethrin with 100% mortality rate: (Fig. 4a).

3.4 Bendiocarb resistance status

Anopheles gambiae s.l. populations showed an increase in Bendiocarb resistance in most of the studied sites, regardless of the cotton growing area, at the scale of study. Only the populations of Koubri (a conventional cotton-growing area) showed a constant susceptibility to Bendiocarb with mortality rates ranging from 98 to 100% in 2008, 2009, 2013 and 2014.
In 2008, resistance was found in seven sites: Banfora, Diébougou, Orodara, Gaoua, Houndé, Soumousso (Conventional cotton growing area) and Dano (Organic cotton growing area) with mortality rates ranged from 29.5 to 82%. During the year 2009, resistance was observed in eleven sites namely Banfora, Orodara, Gaoua, Diebougou, Soumousso, Manga (Conventional cotton growing area), Tiefora, Dano (Biological cotton growing area) and Fada N’Gourma, Koupéla and Manga (Transgenic cotton growing area) with recorded mortality rates varying from 49.5 to 89%.
In 2013, bendiocarb resistance within An. gambiae populations was reported in four sites located in Conventional cotton growing area and one site in Organic cotton growing area where resistance was already observed in both previous years (2008 & 2009). Thus, the mortality rate was 74.9% in Diebougou, 79.7% in Gaoua, 71.4% in Orodara, 79.3% in Soumousso and 76 % in Dano. In 2014, bendiocard resistance was reported in seven study sites: Vallée du Kou, Diebougou, Orodara, Soumousso, Banfora, Tiefora with mortality rates ranged from 64.7 to 79% and more significantly in Banfora and Orodara with very low mortality rate 15.7% and 37.7% respectively.
At study scale, susceptibility seemed partially restored with regard to mortality rate from 2009 to 2014 in Dano (organic cotton area) (62% in 2009; 75% in 2013 and 82 % in 2014). As for Fada N’Gourma in the transgenic area, mortality rate has risen from 90 to 100 %.

3.5 Chlorpyriphos-methyl resistance status

Anopheles gambiae s.l. populations from all study sites were susceptible to Chlorpyriphos-methyl (CM). Overall, mean mortality rates were between 98–100% in the course of the four years (Fig. 6).

3.6 Distribution of Anopheles gambiaecomplex species according to cotton pest control strategy

Overall, identification of sibling species within the An. gambiaecomplex by PCR analysis shown that mosquito populations were composed of a mixture of An. coluzzii, An. gambiae s.s and An. arabiensis with an alike distribution throughout the three settings of cotton cultivation (transgenic, organic and conventional cotton growing area) (Table S3). Anopheles gambiae s.s (55.14%, n=1608/2916) predominated in the three areas of cotton cultivation (transgenic, organic and conventional) following by An. coluzzii(25.06%, n=731/2916) and An. arabiensis (19.78%, n=577/2916) (Table S3).
The statistical analysis did not show any significant difference regarding the distribution of the Anopheles gambiae sl complex species across the three settings of cotton cultivation (Kruskal-wallis, Chi-squared test, \(\chi\)2=1.41, df= 2, p-value = 0.49).

3.7 Impact of cotton pest management strategies on kdr L1014F frequency

The frequencies of kdr L1014F were analyzed according to the members of the An. gambiae complex in the three cotton growing areas (Table S4) and show different patterns of variation from 2008 to 2014 (Fig.7). The three cotton pest control strategies recommended in Burkina Faso did not affect the dynamics of kdrL1014F frequencies in An. gambiae sl populations, nor their evolution over the time of the study (Tukey’s tests, p-values > 0.13).
In the conventional cotton growing area, the frequencies of kdrL1014F mutation were almost close to fixation in An. gambiae sl populations with values reaching 0.95 at most sites located in the western part of the country as Orodara, Banfora, Tiefora and Houndé. However, in Gaoua and Koubri the frequencies were relatively low in An. gambiae s.l in 2008, with values of 0.31 and 0.33 respectively, and then much higher values (0.98) were found in 2014.
.
As for the transgenic cotton growing area, the frequency of thekdr mutation in An. gambiae sl populations was 0.61 in Fada N’Gourma, 0.5 in Koupéla and 0.66 in Manga at the beginning of the study (2008). In 2014, an increase in the kdr frequency was found at these sites with maximum values of 0.98 in both Koupéla and Fada N’Gourma.
In the organic cotton growing area (Tiefora where data were available from 2008 to 2014), the kdr L1014F frequencies were already higher in 2008 with values of 0.92. These frequencies increased in subsequent years to a maximum of 0.98 in 2014.