Effects of Irrigated Rice Fields and Seasonality on Plasmodium Transmission in West Africa , Particularly in Central Côte d ’ Ivoire

The transmission of the parasites that cause human malaria is influenced by myriad environmental factors, including changes in agricultural practices, deforestation, and waterresources development and management (Ijumba and Lindsay, 2001; Ijumba et al., 2002; Keiser et al., 2005; Guerra et al., 2006), climatic factors such as rainfall, humidity and temperature (Reiter, 2008), and various cultural, economic, political and social factors, including health-seeking behaviour, urbanization, armed conflict and war (Esse et al., 2008; Baragatti et al., 2009). Drug resistance in the causative parasites and insecticide resistance in the mosquito vectors are also important factors that now influence malaria transmission (Reiter, 2008). Each year, in many parts of Africa, the local populations of anopheline mosquitoes build up rapidly and peak shortly after the onset of the rainy season (Mbogo et al., 1995). In two studies on the relationships between mosquito abundance, malaria transmission and rainfall in West Africa, 70%–90% of the children investigated were found infected with Plasmodium spp. after the rainy season (Bonnet et al., 2002; Koudou et al., 2009). It is particularly during and at the end of the rainy season that malaria becomes one of the leading causes of mortality and health-seeking at dispensaries and hospitals in this region (Rey et al., 1987). Not only season but also changing patterns of agriculture, particularly irrigated rice farming, influence malaria transmission in Africa (Ijumba & Lindsay, 2001; Diuk-Wasser et al., 2007; Sogoba et al., 2007). Additionally, malaria transmission, Plasmodium prevalence rates, the proportion of presumptive and clinically-confirmed malaria episodes have been studied in two villages of central Côte d’Ivoire: one with irrigated rice farming (Zatta) and one without (Tiémélékro) (Koudou et al., 2009). Due to a farmers’ conflict over land and socio-political issues, irrigated rice farming was interrupted in Zatta in 2003. The goal of this contribution to a book chapter is to evaluate the relationship between Plasmodium transmission, seasonality and agriculture practices.


Introduction
The transmission of the parasites that cause human malaria is influenced by myriad environmental factors, including changes in agricultural practices, deforestation, and waterresources development and management (Ijumba and Lindsay, 2001;Ijumba et al., 2002;Keiser et al., 2005;Guerra et al., 2006), climatic factors such as rainfall, humidity and temperature (Reiter, 2008), and various cultural, economic, political and social factors, including health-seeking behaviour, urbanization, armed conflict and war (Esse et al., 2008;Baragatti et al., 2009).Drug resistance in the causative parasites and insecticide resistance in the mosquito vectors are also important factors that now influence malaria transmission (Reiter, 2008).Each year, in many parts of Africa, the local populations of anopheline mosquitoes build up rapidly and peak shortly after the onset of the rainy season (Mbogo et al., 1995).In two studies on the relationships between mosquito abundance, malaria transmission and rainfall in West Africa, 70%-90% of the children investigated were found infected with Plasmodium spp.after the rainy season (Bonnet et al., 2002;Koudou et al., 2009).It is particularly during and at the end of the rainy season that malaria becomes one of the leading causes of mortality and health-seeking at dispensaries and hospitals in this region (Rey et al., 1987).Not only season but also changing patterns of agriculture, particularly irrigated rice farming, influence malaria transmission in Africa (Ijumba & Lindsay, 2001;Diuk-Wasser et al., 2007;Sogoba et al., 2007).Additionally, malaria transmission, Plasmodium prevalence rates, the proportion of presumptive and clinically-confirmed malaria episodes have been studied in two villages of central Côte d'Ivoire: one with irrigated rice farming (Zatta) and one without (Tiémélékro) (Koudou et al., 2009).Due to a farmers' conflict over land and socio-political issues, irrigated rice farming was interrupted in Zatta in 2003.The goal of this contribution to a book chapter is to evaluate the relationship between Plasmodium transmission, seasonality and agriculture practices.

Study sites
The study described here was carried out in the villages of Tiemelekro (geographical coordinates: 6˚500 N, -4˚170 W) and Zatta (6˚880 N, -5˚390 W), located in central Côte d'Ivoire (Figure 1).A detailed description of Tiemelekro, including climatic conditions, current health care delivery structures and key demographic and socioeconomic indicators, has been presented recently (Girardin et al., 2004).Zatta is located 7 km north-west of Yamoussoukro, the capital city of Côte d'Ivoire.The mean annual temperature in this village is 26.5˚C and the mean annual precipitation is 1280 mm.There is a long rainy season between April and July and a shorter one in October/November.A dispensary, run by two local nurses, is located in Zatta and also covers nearby settlements.Two small dams were constructed in this village in the mid-1970s.Since 1997, a very large irrigated rice field has been cultivated on an estimated surface area of 36 ha, in close proximity to human habitations.However, due to unstable socio-political conditions and a farmers' conflict over land, rice irrigation was interrupted in 2000 and again in 2003/2004.Living conditions and several of the investigated household characteristics are comparable between the two study villages.For example, similar proportions of houses utilized ironcorrugated sheets as roofing material (93.8% in Tiemelekro vs. 92.9% in Zatta), and had running water at home (74.1% vs. 65.4%).On the other hand, improved sanitation facilities were less prominent in Tiemelekro than in Zatta (17.0% vs. 47.6%).With regard to personal protective measures against mosquito bites, the proportion of people sleeping under a bednet was similarly low in both villages (8.4-11.2%),whereas use of fumigating coils was much more pronounced in Zatta (47.3%) when compared to Tiemelekro (9.1%).

Rainfall data collection
The Ivorian "Societe d'Exploitation et de Developpement Aeroportuaire et Meteorologique" (SODEXAM) holds rainfall data for the study area, from 1971 onwards.For the present study, monthly rainfall data from 2002 to 2005 were extracted from the society's records.

Adult mosquitos' collection
Overall, 13 entomological surveys were carried out: seven in the long rainy seasons (in the April and June of 2002, the April, May, June and July of 2003, and the May of 2005), and six in dry seasons (in the February and August of 2002, 2003and 2005).Adult mosquitoes were collected by means of human-bait night catches.The surveys in 2002 and 2005 were each conducted between 18.00 and 06.00 hours, both inside and outside sentinel houses.Each of the six surveys carried out in 2003, however, covered shorter time periods (from 22:00 to 06:00 hours) and the collectors were only stationed inside the sentinel house because of the unstable socio-political situation at the time.Overall, 96, 48 and 32 night catches were carried out in 2002, 2003 and 2005, respectively.No surveys could be undertaken in 2004, as it was then considered too dangerous to reach the study villages.

Laboratory procedures
Adult mosquitoes were brought to a laboratory and processed.Firstly, the physiological age of each female Anopheles and the corresponding parity rate (i.e. the proportion of female mosquitoes that had laid eggs at least once) were determined by dissection of ovaries and

Results
3.1 Species composition of An. gambiae complex and An.funestus group A total of 110 mosquitoes were identified to species level by PCR: 60 from Tiémélékro and 50 from Zatta.Within Anopheles spp.morphologically identified as An.gambiae complex, 100% were An.gambiae s. s.With regard to the An.funestus group, it consisted of 100% An. funestus s. s.

Plasmodium transmission
Comparison between years revealed that the biting rate of An. gambiae s.l. in Zatta decreased several-fold from 49.3 bites per person per night (b/p/n) in 2002 to 7.9 b/p/n in 2003 (likelihood ratio test (LRT=1072.66;P<0.001).In Tiemelekro, the biting rates recorded in 2002 and 2003 remained fairly constant.These observations were paralleled by a marked decrease in the infective rate of An. gambiae s.l. in Zatta (4.6-1.2%), and an increase in Tiemelekro (3.1-7.6%).Meanwhile, the entomological inoculation rate (EIR) of An. gambiae s.l.decreased 21fold in Zatta, from 789 to 38 infective bites per person per year (ib/p/y), whereas it remained high in Tiemelekro (233 vs. 342 ib/p/y).In Zatta, the return to irrigated rice farming in January 2005 was paralleled by a significant increase of the EIR ranging from 38 infective bites per person per year (ib/p/y) in 2003 to 295 ib/p/y in 2005.In Tiémélékro high EIRs were found in 2003 (342 ib/p/y) and 2005 (572 ib/p/y).

Fever cases and asymptomatic carriers, stratified by parasite density
Table 1 shows how many of the children examined with parasitaemia in the 2003 surveys were either asymptomatic carriers or presented with a fever.There was a strong seasonal variation in the proportion of fever cases among individuals with parasitaemia.In Zatta, for example, the proportion of fever cases among Plasmodium-positive individuals was significantly higher towards the end of the rainy season (August) when compared to the dry season (March) (22.1% versus 9.9%; χ 2 = 9.90, df = 1; P = 0.002).In Tiémélékro, considerably higher frequencies of fever cases among Plasmodium-positive individuals were recorded during the peak rainy season in June (27.3%)and towards the end of the rainy season in August (25.5%)when compared to the dry season in March (15.9%;P < 0.05 for both comparisons).In Zatta, all individuals with a high level of parasitaemia (≥ 5000 parasites/μl blood) presented with fever, accounting for a highly significant difference between the proportion of asymptomatic carriers and fever cases in this parasitaemia class (P < 0.001).Similarly, there was a highly significant association between the fever cases and high parasitaemia in the three surveys carried out in 2003 in Tiémélékro (P < 0.05).No statistically significant difference was found in children with lower parasitaemias (1000-5000 parasites/µl of blood), neither in Zatta (March: χ 2 = 1.53; df = 1; P = 0.068 and August: χ 2 = 0.116; df = 1; P = 0.733) nor in Tiémélékro (March: χ 2 = 0.18; df = 1; P = 0.671, June: χ 2 = 2.23; df = 1; P = 0.135 and August: χ 2 = 0.001; df = 1; P = 0.973).

Annual variation of presumptive cases and malaria transmission
In Zatta, 966, 812, 693 and 884 presumptive cases were recorded in 2002, 2003, 2004 and 2005, respectively.The annual number of presumptive malaria cases decreased significantly by 15.1% and 14.7%, respectively, from 2002 to 2003 (IRR = 0.841, P < 0.001) and from 2003 to 2004 (IRR = 0.853, P = 0.002).An opposite trend was observed from 2004 to 2005; the number of presumptive malaria cases increased significantly by 27.5% (IRR = 1.276,P < 0.001).The monthly number of presumptive cases was not related to the monthly number of infective bites per person (IRR = 0.994, P = 0.827).As in the case of Zatta, the monthly number of presumptive cases was not related to the monthly number of infective bites per person (IRR = 1.007;P = 0.776).

Effects of seasonality on Plasmodium Transmission
Tables 2 and 3  (LRT = 0.17, P = 0.879).In Tiémélékro, in 2002 (LRT = 1.84;P = 0.069) and 2005 (LRT = 0.56; P = 0.455), there were no significant differences in An. gambiae s. s. biting rates between the dry and the rainy season.In 2003, the biting rate was significantly higher in the long rainy season (LRT = 3.87, P <0.001).Regarding An. funestus s. s. biting rates, those recorded in the dry season of 2002 (LRT = 6.15) and 2003 (LRT = 4.50) were significantly higher than those recorded in the rainy season (both P <0.001).The difference in the biting rates between the dry and rainy season in 2005 also showed statistical significance (LRT = 3.26; P = 0.031).In brackets are the number of malaria vectors analyzed; LRT (likelihood ratio test) *Vegetable farming is performed intensively with 2 production cycles per year **Vegetable farming is performed intensively with 1 production cycle per year

Discussion
The interruption of irrigated rice farming due to a farmers' dispute over land property rights, coupled with an unstable socio-political situation in the face of the 2002-2004 armed conflict (Betsi et al., 2006;Fürst et al., 2009) offered a unique opportunity to study the dynamics of malaria transmission.Our analyses complement previous publications (Girardin et al., 2004;Koudou et al., 2005Koudou et al., , 2007Koudou et al., , 2009)), now with an explicit focus on the effect of seasonality on malaria transmission under changing agro-ecological conditions.The following points are offered for discussion.Firstly, biting rates of An. gambiae in both villages were usually significantly higher in the rainy season than in the dry season.When irrigated rice farming was interrupted in Zatta in 2003, much lower biting rates were observed than in the preceding year and in 2005, but there were no seasonal differences.Hence, the interruption of irrigated rice farming appeared to have hidden the effect of season on An. gambiae biting rate.These findings are in agreement with previous investigations in the humid savannah of Côte d'Ivoire: in an area characterised by intensive agriculture, the biting rate of An. gambiae increased significantly a few weeks after the beginning of the rainy season, whereas it decreased and became lowest towards the end of the dry season (Doannio et al., 2006).Moreover, the blunting of seasonal differences in biting rates due to changing patterns in irrigated rice farming has been documented previously for the savannahs of Senegal (Faye et al., 1993) and Mali (Dolo et al., 2004).In contrast to An. gambiae with the highest biting rates usually observed in the rainy season, the highest biting rates of An. funestus were consistently recorded in the dry season regardless of the prevailing agricultural activity.Moreover, interruption of irrigated rice farming in Zatta showed no effect.Secondly, with the only exception of a significantly higher infection rate of An. funesuts in Tiémélékro in the rainy season compared to the dry season of 2005, infection rates of both An.gambiae and An.funestus showed no clear seasonal patterns.Different results were reported from Dielmo, a holoendemic area in Senegal, where the infection rate of malaria vectors showed considerable seasonal variation (Fontenille et al., 1997).The observations made in Senegal corroborated previous findings obtained in the savannah area in the north of Côte d'Ivoire (Dossou-Yovo et al., 1995), and other findings documenting a high infection rate of An. funestus at the beginning of the dry season in an irrigated rice area compared to a non-irrigated rice farming area (Dossou-Yovo, 2000).It should be noted, however that the mean annual infection rate of An. gambiae in Zatta was significantly higher when irrigated rice farming was in place (in 2002 and 2005) compared to a year with interrupted irrigated rice farming (Koudou et al., 2005).
Thirdly, the influence of changing patterns of irrigated rice farming on the An.gambiaespecific EIR in Zatta has been discussed elsewhere (Koudou et al., 2005(Koudou et al., , 2007)).In brief, interruption of irrigated rice farming resulted in several-fold lower EIRs compared to normal years.Here, we now document that seasonal patterns of transmission remained.Indeed, considerably higher EIRs were observed for An.gambiae in the rainy season compared to the dry season.Of note, the EIR of An. gambia in the dry season of 2003 in Zatta dropped to zero.In Tiémélékro, high EIRs were recorded throughout the study period for An.gambiae and, in general, EIRs were higher in the rainy season compared to the dry season.An. funestus seemed to play an important role in the transmission of malaria, particularly in the dry season.Our results therefore confirm previous observations made elsewhere in the northern savannah of Côte d'Ivoire (Dossou-Yovo, 2000) and in southern Cameroon (Bonnet et al., 2002).Whilst An. gambiae was the key P. falciparum transmitter mainly during the rainy season, An. funestus was the main vector species during the dry season.It is interesting to note that a previous study focusing on climatic models for suitable malaria transmission in Africa, based on monthly rainfall and temperature data, concluded that an average of 80 mm rainfall per month, for at least 3-5 months, is a minimum to ascertain stable malaria transmission (Craig et al., 1999).Usually, a rapid rise in the An.gambiae population at the beginning of the short rainy season was followed by an increase in the EIR (Bonnet et al., 2002).With regard to An. funestus, the highest EIRs were usually observed during the dry season.Indeed, An. funestus is often abundant and has high EIR during dry season compared to the rainy season (Fontenille et al., 1997;Manga et al., 1997).An. funestus was identified as the main malaria vector in the Guinean climatic region, in East Africa and Madagascar (Robert et al., 1985;Severini et al., 1990).As shown in our study, despite the presence of irrigated rice field, there is a great variability in the annual EIR values and seasonality would seem to play a key role (Mabaso et al., 2007).
Finally, an important finding of our study is that in Zatta, where irrigated rice farming was interrupted in 2003/2004, Plasmodium prevalence rates and the number of presumptive malaria cases decreased.This observation is corroborated by a significant decrease in the EIR from 2002 to 2003 (Koudou et al., 2005) and a significant increase from 2003 to 2005 (Koudou et al., 2007).This study demonstrated also that irrigated rice cultivation is associated with elevated malaria prevalence rates, as well as high numbers of presumptive malaria cases, as seen in Burundi (Coosemans, 1985), Kenya (Githeko et al., 1993) and Madagascar (Marrama et al., 1995).However, research carried out in Tanzania showed that irrigated rice farming was not associated with a higher risk of malaria.One important reason for this observation is that farmers engaged in irrigated rice have the opportunity to gain some extra money, part of which is spent for protective measures against malaria.A reduced risk of malaria despite enhanced rice production has been termed 'paddies paradox' (Ijumba & Lindsay, 2001).

Conclusion
In conclusion, analyses of our entomological data revealed that malaria transmission in two different agro-ecological settings of central Côte d'Ivoire is very high, but there are clear seasonal patterns.Whilst the interruption of irrigated rice farming in one of the two study villages resulted in a highly significant reduction in the EIR, seasonal patterns of transmission remained.Hence, even in intensive agriculture areas, the effect of season on malaria transmission must be taken into consideration for the design of integrated interventions and their monitoring.Additionally, in Zatta, from 2002 to 2003, the highly significant reduction in the annual EIR was paralleled by a significant reduction in the Plasmodium prevalence rate, and the proportions of presumptive and clinically-confirmed malaria cases.Once irrigated rice farming was resumed, there was an increase in entomological and parasitological parameters of malaria.In Tiémélékro, despite the significant increase in the EIR from the year 2002 to 2005 (Koudou et al., 2005(Koudou et al., , 2007)), malaria prevalence rates, and the presumptive and clinical malaria cases decreased.Hence, the reduction of malaria transmission in endemic areas does not necessary reduce the incidence of clinical malaria episodes (Charlwood et al., 1998), highlighting the complex relationship between these parameters.

Acknowledgements
Our thanks are addressed to the local authorities and villagers of Tiémélékro and Zatta, and the district health officers of Dimbokro and Yamoussoukro, for their commitment in the present study.This investigation received financial support from the 'Fonds Ivoiro Suisse de Effects of Irrigated Rice Fields and Seasonality on Plasmodium Transmission in West Africa, Particularly in Central Côte d'Ivoire 159

Table 1 .
Number (%) of children infected with P. falciparum who were asymptomatic carriers or presented with fever, stratified by different levels of parasitaemia, in the two study villages of Tiémélékro and Zatta, central Côte d'Ivoire.Effects of Irrigated Rice Fields and Seasonality on Plasmodium Transmission in West Africa, Particularly in Central Côte d'Ivoire 161In Tiémélékro, the yearly numbers of presumptive malaria cases were 2089, 1858, 1655 and 1541.Thus, we observed significant decreases in the yearly number of presumptive cases by 11.1% from 2002 to 2003 (IRR = 0.889, P < 0.001), 9.0% from 2003 to 2004 (IRR = 0.910, P = 0.005) and 8.9% from 2004 to 2005 (IRR = 0.911, P = 0.008). www.intechopen.com

1 Relationship between season and biting and infection rates
summarise the mean with 2003te, infection rate and the entomological inoculation rate (EIR) of An. gambiae and An.funestus in the two study villages in2002, 2003  and 2005.In Zatta, significantly higher An.gambiae s. s. biting rates were recorded in the dry seasons of 2002 and 2005 when irrigated rice farming was practiced, compared to the dry season of 2003 when irrigated rice farming was interrupted(LRT comparing 2002with 2003:  13.79, LRT comparing 2005 with 2003: 20.50; both P <0.001).In 2003, there was no seasonal difference in the biting rate of An. gambiae s.

Table 2 .
Monthly average biting rate, infection rate, parity rate and entomological inoculation rate (EIR) of An. gambiae and An.funestus during the dry season and the rainy season in 2002, 2003 and 2005 in Zatta, central Côte d'Ivoire www.intechopen.comEffects of Irrigated Rice Fields and Seasonality on Plasmodium Transmission in West Africa, Particularly in Central Côte d'Ivoire 163

Table 3 .
Monthly average biting rate, infection rate, parity rate and entomological inoculation rate (EIR) of An. gambiae and An.funestus during the dry season and the rainy season in 2002, 2003 and 2005 in Tiémélékro, central Côte d'Ivoire Effects of Irrigated Rice Fields and Seasonality on Plasmodium Transmission in West Africa, Particularly in Central Côte d'Ivoire 165 www.intechopen.com