The malaria case incidence has been recorded from 14 RHCs for each calendar week since August 2008 (the second site in the flood plain, Kabulamwanda was incorporated in the study in 2010 and is not shown in Figures
1 and
3). The patterns of transmission in the various topographical zones (Figure
2) are discernible and define the relative importance of the overall topography each season. Two centres are in a flood plain ecosystem, a well-watered, somewhat homogeneous zone of grassland with a fairly high water table approximately 2–4 m deep. There are four RHC in the “transitional” zone that is more elevated in hilly ranges with shrub and small tree vegetation and numerous agricultural lands. The remainder are in the ecological heartland, an area of general agriculture and settlement[
5] but where the water table is some 20 m deep. Data points in Figure
2 are a mean of three annual readings: 2008, 2009 and 2010 and show that there is a regular five-month period (June-October) when the occurrence of cases within the whole study area is confined mainly to the flood plain environment. This is apparently a bottleneck in the parasite population and will perhaps be vulnerable to elimination and can be seen more clearly in Figure
3 (A-D). During the period 2008–9 there was widespread distribution and use of insecticide-treated bed nets which affected transmission of malaria in the heartland and transitional zone. Data in Figure
4 show the result of this intervention in the three topographical zones during 2010, 2011 and 2012. Perusal of Figure
4, where the mean weekly incidence of malaria is shown for weeks 1–20 for 2009, 2010, 2011 and 2012, indicates a discernible difference in the incidence rates between 2009 and 2012 early intervention and three years post intervention, from the 2010 and 2011 seasons when some reduction in transmission was evident. One could postulate this was due to a decline in the insecticidal efficacy of the intervention which commenced in mid-2008[
13]. This effect is seen best from week 1–10 in all three ecological zones. Looking at the patterns of transmission (Figures
2 and
3) during the period weeks 30–46, there was evidence of sustained but low level of transmission throughout the five-month dry and hot period each year in the floodplain only. Sporadic cases occurred in the transition area but almost no cases presented in the clinics of the heartland. The patterns suggest that the cases detected in the flood plain during the low transmission period represent local hot spots or reservoirs, and that these foci maintain the residual parasite population to infect the next generation of mosquitoes which then spread throughout the area.