Large fluctuations in the effective population size of the malaria mosquito Anopheles gambiae s.s. during vector control cycle

Abstract

On Bioko Island, Equatorial Guinea, indoor residual spraying (IRS) has been part of the Bioko Island Malaria Control Project since early 2004. Despite success in reducing childhood infections, areas of high transmission remain on the island. We therefore examined fluctuations in the effective population size (N e ) of the malaria vector Anopheles gambiae in an area of persistent high transmission over two spray rounds. We analyzed data for 13 microsatellite loci from 791 An. gambiae specimens collected at six time points in 2009 and 2010 and reconstructed the demographic history of the population during this period using approximate Bayesian computation (ABC). Our analysis shows that IRS rounds have a large impact on N e , reducing it by 65%-92% from prespray round N e . More importantly, our analysis shows that after 3-5 months, the An. gambiae population rebounded by 2818% compared shortly following the spray round. Our study underscores the importance of adequate spray round frequency to provide continuous suppression of mosquito populations and that increased spray round frequency should substantially improve the efficacy of IRS campaigns. It also demonstrates the ability of ABC to reconstruct a detailed demographic history across only a few tens of generations in a large population.

Keywords: Anopheles gambiae; approximate Bayesian computation; effective population size; indoor residual spraying; malaria; vector control.

Figure 1

A map of Bioko Island indicating the location of our sampling site, Mongola, within the Punta Europa area.

Figure 2

A violin plot for N_e estimates for Anopheles gambiae (M) in Punta Europa, Bioko Island, for five intervals during 2009 and early 2010. Horizontal bars indicate the median of the posterior density distribution of N_e, and vertical bars indicate the 95% credibility intervals (Cr. I).

Figure 3

N_e estimates of Anopheles gambiae (M) in Punta Europa, Bioko Island (bars), and monthly precipitation amount (line). Actual rainfall data were collected for the period from March to November 2009 in Punta Europa. Monthly averages from the years 2004–2007 were used for the months for which no 2009 data were available. The dashed lines show the approximate timing of the spray rounds in June 2009 and December 2009.

Figure 4

The effect of large egg–larval–pupal mortality on Anopheles gambiae effective population size as indicated by our post hoc demographic model. Two events were modeled, a 99% and 75% mortality of eggs–larvae–pupae for a single day.

Figure 5

A visualization of the fluctuations in N_e in Anopheles gambiae prior and post-IRS interventions based on the results from Athrey et al. ; and this study. Before the IRS campaign, seasonal fluctuations existed within the populations. After the IRS campaign started, N_e was reduced but continues to fluctuate, now largely due to the timing of IRS rounds.