Robert J. Novak

We examined algorithms for malaria mapping using the impact of reflectance calibration uncertainties on the accuracies of three vegetation indices (VI)'s derived from QuickBird data in three rice agro-village complexes Mwea, Kenya. We also generated inferential statistics from field sampled vegetation covariates for identifying riceland Anopheles arabiensis during the crop season. All aquatic habitats in the study sites were stratified based on levels of rice stages; flooded, land preparation, post-transplanting, tillering, flowering/maturation and post-harvest/fallow. A set of uncertainty propagation equations were designed to model the propagation of calibration uncertainties using the red channel (band 3: 0.63 to 0.69 μm) and the near infra-red (NIR) channel (band 4: 0.76 to 0.90 μm) to generate the Normalized Difference Vegetation Index (NDVI) and the Soil Adjusted Vegetation Index (SAVI). The Atmospheric Resistant Vegetation Index (ARVI) was also evaluated incorporating the QuickBird blue band (Band 1: 0.45 to 0.52 μm) to normalize atmospheric effects. In order to determine local clustering of riceland habitats Gi*(d) statistics were generated from the ground-based and remotely-sensed ecological databases. Additionally, all riceland habitats were visually examined using the spectral reflectance of vegetation land cover for identification of highly productive riceland Anopheles oviposition sites.

The survivorship and distribution of Anopheles arabiensis larvae and pupae was examined in a rice agro-ecosystem in Mwea Irrigation Scheme, central Kenya, from August 2005 to April 2006, prior to implementation of larval control programme. Horizontal life tables were constructed for immatures in semi-field condition. The time spent in the various immature stages was determined and survival established. Vertical life tables were obtained from five paddies sampled by standard dipping technique. Pre-adult developmental time for An. arabiensis in the trays in the experimental set up in the screen house was 11.85 days from eclosion to emergence. The mean duration of each instar stage was estimated to be 1.40 days for first instars, 2.90 days for second instars, 1.85 days for third instars, 3.80 days for fourth instars and 1.90 days for pupae. A total of 590 individuals emerged into adults, giving an overall survivorship from L1 to adult emergence of 69.4%. A total of 4,956 An. arabiensis immatures were collected in 1,400 dips throughout the sampling period. Of these, 55.9% were collected during the tillering stage, 42.5% during the transplanting period and 1.6% during the land preparation stage. There was a significant difference in the An. arabiensis larval densities among the five stages. Also there was significant variation in immature stage composition for each day's collection in each paddy. These results indicate that the survival of the immatures was higher in some paddies than others. The mortality rate during the transplanting was 99.9% and at tillering was 96.6%, while the overall mortality was 98.3%. The survival of An. arabiensis immatures was better during the tillering stage of rice growth. Further the survival of immatures in rice fields is influenced by the rice agronomic activities including addition of nitrogenous fertilizers and pesticides. For effective integrated vector management, the application of larvicides should target An. arabiensis larvae at the tillering stage (early vegetative stage of rice) when their survival in the aquatic habitats is high to significantly reduce them and the larvicides should be long-lasting to have a significant impact on the malaria vector productivity on the habitats.

For remote identification of mosquito habitats the first step is often to construct a discrete tessellation of the region. In applications where complex geometries do not need to be represented such as urban habitats, regular orthogonal grids are constructed in GIS and overlaid on satellite images. However, rice land vector mosquito aquatic habitats are rarely uniform in space or character. An orthogonal grid overlaid on satellite data of rice-land areas may fail to capture physical or man-made structures, i.e paddies, canals, berms at these habitats. Unlike an orthogonal grid, digitizing each habitat converts a polygon into a grid cell, which may conform to rice-land habitat boundaries. This research illustrates the application of a random sampling methodology, comparing an orthogonal and a digitized grid for assessment of rice land habitats. A land cover map was generated in Erdas Imagine V8.7^® using QuickBird data acquired July 2005, for three villages within the Mwea Rice Scheme, Kenya. An orthogonal grid was overlaid on the images. In the digitized dataset, each habitat was traced in Arc Info 9.1^®. All habitats in each study site were stratified based on levels of rice stage The orthogonal grid did not identify any habitat while the digitized grid identified every habitat by strata and study site. An analysis of variance test indicated the relative abundance of An. arabiensis at the three study sites to be significantly higher during the post-transplanting stage of the rice cycle. Regions of higher Anopheles abundance, based on digitized grid cell information probably reflect underlying differences in abundance of mosquito habitats in a rice land environment, which is where limited control resources could be concentrated to reduce vector abundance.

Continuous land cover modification is an important part of spatial epidemiology because it can help identify environmental factors and Culex mosquitoes associated with arbovirus transmission and thus guide control intervention. The aim of this study was to determine whether remotely sensed data could be used to identify rice-related Culex quinquefasciatus breeding habitats in three rice-villages within the Mwea Rice Scheme, Kenya. We examined whether a land use land cover (LULC) classification based on two scenes, IKONOS at 4 m and Landsat Thematic Mapper at 30 m could be used to map different land uses and rice planted at different times (cohorts), and to infer which LULC change were correlated to high density Cx. quinquefasciatus aquatic habitats. We performed a maximum likelihood unsupervised classification in Erdas Imagine V8.7® and generated three land cover classifications, rice field, fallow and built environment. Differentially corrected global positioning systems (DGPS) ground coordinates of Cx. quinquefasciatus aquatic habitats were overlaid onto the LULC maps generated in ArcInfo 9.1®. Grid cells were stratified by levels of irrigation (well-irrigated and poorly-irrigated) and varied according to size of the paddy. Total LULC change between 1988–2005 was 42.1 % in Kangichiri, 52.8 % in Kiuria and and 50.6 % Rurumi. The most frequent LULC changes was rice field to fallow and fallow to rice field. The proportion of aquatic habitats positive for Culex larvae in LULC change sites was 77.5% in Kangichiri, 72.9% in Kiuria and 73.7% in Rurumi. Poorly – irrigated grid cells displayed 63.3% of aquatic habitats among all LULC change sites. We demonstrate that optical remote sensing can identify rice cultivation LULC sites associated with high Culex oviposition. We argue that the regions of higher Culex abundance based on oviposition surveillance sites reflect underlying differences in abundance of larval habitats which is where limited control resources could be concentrated to reduce vector larval abundance.

Entomological studies were conducted over a 24-month period in 8 villages to establish the behavior patterns, seasonal densities, and variation in entomological inoculation rates (EIRs) of Anopheles arabiensis, the main vector of malaria in Eritrea. A total of 5,683 anopheline mosquitoes were collected through indoor sampling (1,613), human-landing catches (2,711), and outdoor pit shelters (1,359). Overall, An. arabiensis was the predominant species at all of the study sites, with its population density increasing during the rainy season. Peak indoor-resting densities was observed during September and October. Human landing indices for An. arabiensis averaged 1.9 and 3.8 per person per night in October and September, respectively. Peak biting and landing rates occurred between 2000-2200 h and 0100-0300 h. Of the total number of bites, 44.7% occurred between 1800 and 2300 h, and at least 56.5% of the total bites occurred outdoors, indicating that the species was partially exophagic. The fed to gravid ratio for An. arabiensis in indoor-resting collections was 2:1, indicating some degree of exophily. The sporozoite rates (SRs) for An. arabiensis ranged from 0.54% in the Anseba zone to 1.3% in the Gash-Barka zone. One mosquito each of An. d'thali (SR = 0.45%) and An. cinereus (SR = 2.13%) was found to be positive. Of the total positive An. arabiensis (n = 64), 18.2% came from human-landing collections outdoors. Blood-meal analysis by enzyme-linked immunosorbent assay for An. arabiensis indicated that this species was partially zoophilic with a human to bovine ratio of 2:1 being recorded. The EIR profiles indicated that malaria transmission is highly seasonal, increasing during the wet season and declining drastically during the dry season. On average, the greatest risk of infection occurs in Hiletsidi, in the Gash-Barka zone (6.5 infective bites per month). The exophilic behavior and early evening biting of An. arabiensis present obstacles for control with treated bed-nets and indoor residual spraying within the context of integrated malaria control, and call for greater focus on strategies such as larval control.