Introduction
World Health Organization estimates that malaria is responsible for about 655,000 deaths annually, with P. falciparum as the leading causative agent (Gathany, 2012). Malaria is also an important disease in foci located in irrigation-based agricultural areas where Anopheles gambiae Giles, 1902 and Anopheles arabiensis Patton, 1905 mosquitoes are predominant vectors. This is because they can readily proliferate due to the abundance of paddies (Mwangangi et al., 2007). Integrated Malaria Management (IMM) is one of the vector borne disease management strategies encompassing the use of insecticides, environmental modifications to discourage mosquito breeding, health education and chemotherapy. This strategy has significantly reduced malaria transmission and burden on these foci (Dolo et al., 2004). The important principle with this strategy is the collective contribution of the individual components which has contributed to the overall reduction in the malaria burden.

However, the success of the IMM program has some challenges, especially with use of insecticides. This is because, resistance to conventional insecticides, a phenomenon currently widespread across Sub-Sahara Africa has been reported (Ochomo et al., 2014; Russell et al., 2010). Resistance to pyrethroids presents a real and immediate challenge to the efficacy of otherwise successful insecticide treated nets (ITN) based malaria control intervention against adult vector (Russell et al., 2010). The long-lasting insecticidal nets (LLIN) and/ or indoor residual spraying (IRS) has influenced An. gambiae population and their host seeking behavior with demographic coverage of humans dramatically reducing in relation to infection prevalence of Plasmodia in An. gambiae (Russell et al., 2010; 2011). 

Previous studies in Mwea and Ahero (Kamau and Vulule, 2006; Kamau et al., 2007) found no resistance to insecticide groups that have been recommended by WHO for Indoor Residual Spraying, including pyrethroids. Prolonged continuous use of insecticides such as that which has been carried out in the country by the Ministry of Health has been shown to result in increased resistance in various settings (Jeffrey and Pia, 2002). Developments that could impact insecticide resistance in Kenya include increased awareness of the effectiveness of ITNs as a result of concerted social marketing efforts such as was witnessed between 2002 and 2004 and increased availability of ITNs either through the provision of heavily subsidized ITN distribution by the Ministry of Health or through free mass ITN distribution through various programmes. Additionally, the use of pesticides for agriculture has also been found to be key driver of resistance in mosquito populations (Riveron and Wondji, 2014). It is widely accepted that increases in insecticide resistance will negatively impact the sustainability of vector control strategies employing insecticides. Thus, continued monitoring of resistance is necessary so as to detect and manage any developing resistance in a timely manner.  The study reported here sought to determine the status of insecticide resistance in Mwea and Ahero some eight to ten years after earlier studies indicated the absence of insecticide resistance and whether there have been changes in blood feeding behavior and malaria parasite rates in the vectors of malaria.

Materials and Methods
Study Sites
This study was conducted between the months of October 2009 and June 2011 in Mwea and Ahero rice irrigation zones. Mwea (37.250 E, 1.420 S) is located on the base of Mount Kenya, Central Kenya, with a population of about 160,000 (Kenya National Bureau of Statistics, 2009). Mwea experiences long and short rain seasons from March to June and October to December, respectively. The mean annual rainfall level is between 1200 to 1600 mm with rice cultivation as the main agro-economic activity. Rice cultivation involves flooding of rice paddies thus potential breeding sites for mosquitoes. The species of Anopheles mosquitoes in the Mwea region include Anopheles arabiensis Patton 1905, Anopheles funestus Giles (Kamau and Vulule, 2006; Ijumba et al., 2008; Muturi et al., 2008) and Anopheles gambiae sensu stricto Giles 1902 (Ijumba et al., 2008; Kamau et al., 2007). Recent studies in Mwea have reported susceptibility of An. gambiae to the conventional insecticides, especially permethrin and deltamethrin (Kamau and Vulule, 2006).

Ahero (34.900 E, 0.160 S) is located 24 Km East of Kisumu town along the shores of Lake Victoria in Western Kenya. It has a human population of 50, 730 (Kenya county fact sheets December 2011). The site experiences long and short rainy seasons from March to May and September to December, respectively. The mean annual rainfall level is between 1,000 to 1,800 mm with rice and sugar cane cultivation and fishing forming part of their major agro-economic activities. In Ahero, the presence of river Nyando facilitates rice fields’ irrigation thus breeding grounds for mosquitoes. The species of Anopheles mosquitoes in Ahero include Anopheles arabiensis Patton 1905, Anopheles funestus Giles (Bruhnes, 1978) and Anopheles gambiae sensu stricto Giles 1902 (Chandler and Highton 2009). Reports on insecticides resistance indicate An. gambiae is susceptible to conventional insecticides (Chandler and Highton 2009). Apart from feeding on humans, non- human hosts include bovines, goats, dogs, felines, birds and reptiles (Githeko et al., 1994).

Sampling, Rearing and Identification of adult Anopheles mosquitoes
Three villages (approximately 20 Km apart) were selected and sampled from each study site. That is, Mbuinjeru, Ndindiruku and Murinduko villages in Mwea scheme; Kamagaga, Kobura and Wagai villages in Ahero scheme. Indoor sampling was carried out as described by Mathenge et al., (2004); Kamau et al., (2007). Briefly, 15 houses were randomly sampled for gravid and blood fed female Anopheles mosquitoes and a total of 200 mosquitoes from each village were collected. This was conducted between 0600Hrs and 0800Hrs. Outdoor sampling was done overnight using CDC light traps model 512 (John W. Hock Company, Gainesville, Fl, USA) with each trap placed a meter from respective households after which sorting was done for gravid and blood fed Anopheles mosquitoes.

The mosquitoes were morphological classified in situ into An. gambiae s.l or An. funestus complex by taxonomic methods of Gillies and De Meillon (1968) and kept separately in paper cups. The mosquitoes were maintained live and subsequently transferred to insectary where standard procedures for rearing Anopheles mosquitoes were followed. Individual females were subsequently sorted based on their physiological status. Individual gravid females were placed in tubes and allowed to oviposit and F₁ families raised separately. After emergency, adult mosquitoes were provided with 10% sugar solution soaked in cotton wool for 2-3 days before conducting bioassays. In cases where F₁ was less than 100, that is, in Kamagaga and Wagai for An. arabiensis and Kobura for An. funestus, F₂ was used for the bioassays. After oviposition, the field collected mosquitoes were used to determine which member of the species complex they belong to, thus their offspring. DNA was extracted by alcohol precipitation method of Collins et al., (1987) and the species-specific analysis done using PCR method of Scott et al., (1993).

Insecticide resistance testing
Bioassays were conducted using 2-3 day old mosquitoes for insecticides in each of the four classes of insecticides recommended by WHO; pyrethroids (permethrin and deltamethrin), organochlorides (DDT), carbamates (bendiocarb) and organophosphates (fenitrothion). Assessment was done using papers impregnated with 0.75 permethrin, 0.05 deltamethrin, 0.04 DDT, 0.01bendiocarb or 1.00% fenitrothion insecticides following WHO guidelines (WHO, 2005). Briefly, 20-25 mosquitoes (four replicates each) were exposed to insecticide-treated papers for an hour (permethrin, bendiocarb or DDT) or two hours (fenitrothion). An. gambiae Kisumu susceptible strain (KSM Strain) was used as positive control while negative control comprised the F₁ mosquitoes exposed to paper treated only with silicon oil solvent. Knock-down (KD) counts were recorded in every 10 minutes. After the exposure period, the mosquitoes were transferred into individual holding tubes without insecticide treated papers and provided with 10% sugar solution and mortality recorded 24 hours later.

Determination of Blood meal sources and Plasmodium falciparum infection
Mosquito abdomens with visible blood meals were individually tested for the source of the blood according to Beier et al. (1988). A total of 405 female An. arabiensis from Mwea study site were assayed. In Ahero, a total of 169 An. arabiensis and 269 An. funestus abdomens were assayed for the sources of the blood meal. A total of 1200 heads and thoraces of all the collected mosquitoes from both study sites (after oviposition and stunning at 4˚C and abdomen removal) were tested for the presence of P. falciparum by sandwich Enzyme Linked Immunosorbent Assay (ELISA) method of Wirtz et al., (1987).

Results
A total of 600 female Anopheles gambiae s.l mosquitoes were collected (355 indoor and 245 outdoor) from Mwea sentinel site as shown in Table 1. All the Anopheles gambiae s.l was identified as An. arabiensis by PCR (Table 2). In Ahero, An. gambiae s.l was 250 (115 indoor and 135 outdoor) and An. funestus was 350 (195 indoor and 155 outdoor) as shown in Table 1. All the An. gambiae s.l were identified as An. arabiensis while An. funestus members were Anopheles funestus sensu stricto Giles, 1900 (342), Anopheles rivulorum Leesoni, 1935 (4), Anopheles leesoni Evans, 1931 (2) and Anopheles parensis Gillies, 1935 (2) as shown in Table 2.

Table 1 Number of Anopheles mosquitoes collected per study site and their species

Table 2 Molecular identification of Anopheles species collected in Mwea and Ahero sentinel sites 2 (a) Anopheles gambiae s.l  2 (b) Anopheles funestus complex

A total of 100 (four replicates of 25), 2-3 days old females of each identified Anopheles species were assayed against each insecticide as shown in Tables 3. In Mwea, mortality 24 h post-exposure was below the 98% threshold for susceptibility in all the assays except for mosquitoes from Mbuinjeru for the test with DDT (98%) and those from Murinduko for the test with Bendiocarb (100%). In Ahero, both permethrin and bendiocarb had a 96% 24 hr post-exposure effect on An. funestus from Kamagaga village. Bendiocarb and fenitrothion demonstrated a 96% and a 98% 24 hr post-exposure effect on An. funestus from Kobura village. Fenitrothion indicated reduced activity on An. arabiensis (96%) and An. funestus (98%) from Wagai village as shown in Table 3.

Table 3 Percentage mortality, KD50 and KD95 and (in bracket) total number of mosquitoes tested against the different insecticides

The Human Blood Index for An. arabiensis sampled from Mwea scheme was at 0.22 (n=405). In Ahero scheme, the Human Blood Index was at 0.17 (n=169) for An. funestus. It was noted that in both study sites, bovine was the preferred host with mixed blood sources. In Mwea, An. arabiensis sourced their blood from humans and animals. In Ahero, An. funestus contacted both humans and animals for blood sources while An. arabiensis sourced their blood from only animals (Table 4).

Table 4 Source of Blood Meals in Anopheles mosquitoes collected in Mwea and Ahero study sites

In Mwea, the entire 600 field collected female An. arabiensis were tested for P. falciparum circumsporozoite infection. None of the samples from Mbuinjeru and Ndindiruku villages tested positive for the parasite infection. In Murinduko, three (3) (n=200) of the samples collected indoor tested positive for the parasite (Table 5). In Ahero study site, Kamagaga village had seven 7 (n=147; 3 indoor, 4 outdoor) of An. funestus positive for the parasite and 4 (n=183; 1 indoor, 3 outdoor) of An. funestus positive for the parasite in Wagai village as shown in Table 5.
 

Table 5 Plasmodium falciparum infection in Anopheles mosquitoes in Mwea and Ahero rice study sites

Discussion
The present study sought to determine the level of insecticide resistance; blood feeding patterns and malaria parasite rates in two areas of Kenya that are endemic for malaria. Overall, levels of resistance against insecticides had increased and malaria parasite rates were lower than previously recorded in both Mwea and Ahero.

In contrast to earlier studies which reported the presence of An. arabiensis Patton and An. funestus Giles in Mwea (Ijumba et al., 2008), the current study found only An. arabiensis. This may be attributed to massive distribution of insecticides treated mosquito nets in these rice schemes thus promoting selection against the population of An. funestus which is known to be anthropophilic and endophagic. In Ahero, the present study identified An. funestus Giles and An. arabiensis Patton as the major Anopheles species. However, this varies from results reported by Chandler et al., (1975) where An. gambiae s.l. was present in a higher proportion than members of An. funestus complex. This variation may be attributed to a gradual change in malaria vector composition since 1975 with unclear causes (Githeko et al., 2003). Anopheles rivulorum, An. leesoni and An. parensis, which are members of the An. funestus complex, had not been reported in Ahero in earlier studies, mostly likely due to the lack of molecular assays to differentiate between members of the species complex.

In terms of resistance to insecticides, there was evidence of reduced susceptibility of An. arabiensis to insecticides in Mwea, especially to pyrethroids when compared to studies of Kamau and Vulule (2006). This indicates that resistance is spreading and is likely to increase further with the continued use of insecticide-based vector interventions. An. arabiensis sampled from Ahero were susceptible to the test insecticides except for fenitrothion which indicated reduced activity. Similarly, the use of permethrin and bendiocarb in Ahero should be monitored as reduced insecticidal activity has been demonstrated in Kamagaga and Kobura villages. Therefore, insecticides based malaria control programs can be effectively used in this ecosystem with close monitoring for resistance development or in integrated vector control programs.

In Mwea, only mosquitoes collected from one of the three villages were positive for P. falciparum. In the said village, Murinduko, malaria parasite rates were much lower than those reported earlier, which were three-fold higher (Muturi et al., 2008). The lower malaria parasite rates in the current study are likely due to the use of bed nets as reported by Kamau and Vulule (2006) in which 93% of the 42 households they surveyed used bed nets and also due to high zoophilic behaviour of An. arabiensis (Muturi et al., 2009). In Ahero, only members of An. funestus complex were reported to be positive for P. falciparum. This indicates the potential of these members to transmit malaria parasites in this region.

In Mwea study site, assessment on the feeding pattern indicated mixed feeding for a blood meal. These findings compare well with those published earlier by Muturi et al., (2008). This phenomenon can be utilised by individuals to keep animals around their homesteads while themselves sleep under a treated mosquito net. This is because; the vector in this site can still obtain a blood meal from animals, especially bovines, as the preferred host, and this will reduce malaria parasite transmission to humans.

In Ahero, we found lower human blood indices for both An. arabiensis and An. funestus mosquitoes sampled compared to those reported by Githeko et al., (1994). The entire mosquito samples from Kamagaga showed animal preference. This should be further investigated to determine if it is due to human inaccessibility or intrinsic behaviour of the vectors. In Kobura and Wagai villages, An. funestus obtained their blood meal from both humans and animals thus a potential vector for malaria parasite while An. arabiensis from the same villages obtained their blood meal from none of the test hosts. This calls for investigations to determine the source of blood meal by An. arabiensis in these villages.

Feeding on multiple hosts, including non-human hosts in Anopheles populations, has been suggested as one of the factors responsible for lower levels of malaria transmission despite higher vector densities (Lincithicum et al., 1999; Maria et al., 2005). However, because An. funestus s.s is substantially more anthropophagic and endophagic than An. arabiensis, this species may not be a good candidate of zooprophylaxis.

In conclusion, the current study found increased levels of resistance to insecticides in both Mwea and Ahero, although resistance was much higher in Mwea.  The lower malaria parasite rates observed in both Mwea and Ahero are consistent with the expected effectiveness of insecticide-based intervention against malaria and our results strongly suggest that current levels of resistance have not compromised their effectiveness.  While epidemiological studies will be needed to determine what resistance levels interfere with the effectiveness of such interventions, the need for continued monitoring of resistance so as to detect and manage it cannot be overstated.

Competing interests:
The authors declare no competing interests.

Authors contributions
JC is the corresponding author and the main researcher of this work.
POM participated in the critical revision of this work and guidance in the molecular analysis.
JNM assisted in the care and use of animals in the laboratory.
LK supervised field visits, all laboratory experiments and rearing of the Anopheles mosquitoes.
CM was the overall supervisor for the whole work right from its funding to conclusion.

Acknowledgements
This work was funded by the Bill and Melinda Gates Foundation and the World Health Organization (Grant No. 1580) and is published with the permission of the Director, Kenya Medical Research Institute (KEMRI). Authors are grateful to villagers and households owners in Mwea and Ahero for consent to mosquito collection from their homesteads.

References
Beier J.C., Perkins P.V., Wirtz R.A., Koros J., Dggs D., Gargan T.P., and Koech D.K.,1988, Blood meal identification by direct Enzyme Linked Immunosorbent Assay (ELISA), tested on Anopheles (Diptera: Culicidae) in Kenya, Medical Entomology, 25: 9-16
http://dx.doi.org/10.1093/jmedent/25.1.9

Chandler J.A., and Highton R.B., 2009, The succession of Mosquito species (Diptera, Culicidae) in rice fields in the Kisumu area of Kenya, and their possible control, Bulletin of Entomology Resources, 65: 295-302 
http://dx.doi.org/10.1017/S0007485300005988

Chandler J.A., Highton R.B., and Hill M.N., 1975, Mosquitoes of the Kano plain, Kenya, Medical Entomology, 12: 504-510
http://dx.doi.org/10.1093/jmedent/12.5.504

Collins F.H., Mendez M.A., Rasmussen M.O., Mehaffey P.C., Besansky N.J., and Finnerty V., 1987, A ribosomal RNA gene probe differentiates member species of the Anopheles gambiae complex, Tropical Medicine and Hygiene, 37:37-41

Dolo G., Briet O.J., Dao A., Traore S.F., Bouare M., Sogoba N., Niare O., Bagayogo M., Sangare D., Teuscher T. and Toure Y.T., 2004, Malaria transmission in relation to rice cultivation in the irrigated Sahel of Mali, Acta Tropica, 89: 147-159
http://dx.doi.org/10.1016/j.actatropica.2003.10.014

Gathany J., 2012, Insecticides resistance survey in Uganda, Nature, 485: 293

Gillies M.T., and DeMeillon B., 1968, The Anophelinae of Africa South of the Sahara (Ethiopian Zoogeographical Region), Publications of the South African Institute for Medical Research, 54: 131-208

Githeko A.K., Service M.W., Mbogo C.M., Atieli F.A., and Juma F.O., 2003, Sampling of Anopheles arabiensis, Anopheles gambiae sensu lato and Anopheles funestus (Diptera: Culicidae) with CDC light-traps near a rice irrigation area and a sugarcane belt in western Kenya, Bulletin of Entomological Research, 84: 319-324
http://dx.doi.org/10.1017/S0007485300032430

Githeko A.K., Service M.W., Mbogo C.M., Atieli F.K., and Juma F.O., 1994, Origin of blood meal in indoor and outdoor resting malaria vectors in Western Kenya, Acta Tropica, 58: 307-316
http://dx.doi.org/10.1016/0001-706X(94)90024-8

Ijumba J.N., Mwangi R.W., and Beier J.C., 2008, Malaria transmission potential of Anopheles mosquitoes in the Mwea-Tebere irrigation scheme in Kenya, Medical and Veterinary Entomology, 4: 425-432
http://dx.doi.org/10.1111/j.1365-2915.1990.tb00461.x

Jeffrey S., and Pia M., 2002, The economic and social burden of malaria, Nature, 415: 680-685
 http://dx.doi.org/10.1038/415680a

Kamau L., and Vulule J., 2006, Status of insecticide susceptibility in Anopheles arabiensis from Mwea rice irrigation scheme, Central Kenya, Malaria, 5: 1475-2874

Kamau L., Agai D., Matoke D., Wachira L., Gikandi G., and Vulule J. M., 2007, Status of insecticide susceptibility in Anopheles gambiae sensu lato and Anopheles funestus mosquitoes from Western Kenya, Insect Science, 8: 1536-2442

Lincithicum K., Anyamba A., Tucket C., Kelly P., Myers M., and Peters C., 1999, Climate and satellite indicators to forecast Rift Valley Fever epidemics in Kenya, Science, 285: 397-400
http://dx.doi.org/10.1126/science.285.5426.397

Maria A., Diuk W., Mahamoudou B.T., Dolo G., Magaran B., Nafoman S., Sekou F.T., Nicholas M., and Charles, E.T., 2005,Vector abundance and malaria transmission in rice-growing villages in Mali, Tropical medicine and Hygiene, 72: 725-731

Mathenge E.M., Omweri G.O., Irungu L.W., Ndegwa P.N., Walczak E., Smith T.A., Killeen G.F., and Knols B.G., 2004, Comparative field evaluation of the Mbita trap, CDC light trap and the human landing catch for sampling of malaria vectors in western Kenya, Tropical Medicine and Hygiene, 70: 33-37

Muturi E.J., Muriu S., Shililu J., Mwangangi J., Jacob B.G., Mbogo C.M., Githure J., and Novak R.J., 2008, Effect of rice cultivation on malaria transmission in central Kenya, American Tropical Medicine and Hygiene, 78: 270-275

Muturi J.E., Kamau L., Benjamin G.J., Muriu S., Mbogo C.M., Shililu J., Githure J., and Novak J.R., 2009, Spatial distribution, blood feeding pattern, and role of Anopheles funestus Complex in malaria transmission in central Kenya, Parasitology research, 105: 1041-1046
http://dx.doi.org/10.1007/s00436-009-1543-5

Mwangangi J.M., Mbogo C.M., Muthuri E.J., Nzovu J.S., Githure J.I., Yan G., Minakwa N., Novak J.R., and Beir J.C., 2007, Spartial distribution and habitation characterization of Anopheles larvae along the Kenyan coast, Vector borne Diseases, 44: 44-51

Ochomo E., Nabie M.B., Luna K., Francis A., John V., Collins O., Maurice O., Kiambo N., David S., Evan M., Lawrence M., Teresa K., Krishanthi S., Immo K., and Charles M., 2014, Pyrethroid susceptibility of malaria vectors in four Districts of western Kenya, Parasitology of Vectors, 47: 310
http://dx.doi.org/10.1186/1756-3305-7-310

Riveron J.M., and Wondji C.S., 2014, A single mutation in the GSTe2 gene allows tracking of metabolically-based insecticide resistance in a major malaria vector, Genome Biology.
http://dx.doi.org/10.1186/gb-2014-15-2-r27

Russell T., Govella N., Azizi S., Drakeley C., and Kachur S.P., 2011, Increased proportions of outdoor feeding among residual malaria vector populations following increased use of insecticide-treated nets in rural Tanzania, Malaria, 10: 80
http://dx.doi.org/10.1186/1475-2875-10-80

Russell T., Lwetoijera D., Maliti D., Chipwaza B., and Kihonda., 2010, Impact of promoting longer-lasting insecticide treatment of bed nets upon malaria transmission in a rural Tanzanian setting with pre-existing high coverage of untreated nets, Malaria, 10: 9: 62

Scott J.A., Brogdon W.G., and Collins, F.H., 1993, Identification of single specimen of Anopheles gambiae complex by the Polymerase Chain Reaction, American Journal of Tropical Medicine and Hygiene, 49: 520-529

Weirtz R., Zavala F., Graves P., and Andre G., 1987, Field evaluation of ELISA for Plasmodium falciparum and Plasmodium vivax sporozoites in mosquitoes (Diptera: Culicidae) from Papua New Guinea, Medical Entomology, 24: 433-437
http://dx.doi.org/10.1093/jmedent/24.4.433

World health organization. 2005, Atlas on insecticide resistance in malaria vectors of the WHO African region, Bulletin of World Health Organization, 78: 5-12