Figure 1.

Intensive Microscopy-Based Quantification of Gametocyte Sex Ratio. In this figure, we use data generated by one of the authors (C.D.) in 1992–1993 to illustrate the likely error in gametocyte sex ratio estimation when determining the sex of only a few gametocytes per sample as routinely done in epidemiological studies. Briefly, 100–500 gametocytes were sexed per blood smear in 53 samples from 43 individuals living in malaria-endemic regions in The Gambia, and sex-specific counts were recorded in groups of 10 gametocytes. Light microscopy images of male and female gametocytes stained with Giemsa’s stain are shown in (A) for illustration. Female gametocytes are blue/violet (as opposed to pink males), more crescent-shaped, and have more compact nuclei and more centrally located pigment. In (B), x coordinates represent data from different samples, partially transparent to allow visualization of overlapping data points: red crosses (left y axis) correspond to the proportions of gametocytes identified as male-based on data from groups of 10 gametocytes or the total number of gametocytes, respectively; light blue bars represent the difference (right y axis) between the proportion of male gametocytes when considering only the first 10 gametocytes observed in each smear versus the proportion calculated based on the total number of gametocytes. This difference exemplifies the error that might occur in routine measurements that quantify only a limited number of gametocytes. In (C), the distribution of the proportion of male gametocytes in the different thick smears analyzed, based on the total number of gametocytes, is shown. Panel (D) presents the progressive reduction in error as the number of gametocytes counted increases. In this panel, the x axis corresponds to the cumulative number of gametocytes sexed, and the y axis corresponds to the difference in the proportion of male gametocytes relative to the same proportion when estimated based on all gametocytes observed in the smear. Each line represents a different sample, and colors relate to the overall proportion of male gametocytes in the smear. The rectangle delimited by the dashed lines encloses error values between −0.1 and 0.1 when 50–100 gametocytes were counted. Only data from thick smears were used in this figure.