Analysing the oviposition behaviour of malaria mosquitoes: design considerations for improving two-choice egg count experiments
© Okal et al. 2015
Received: 15 January 2015
Accepted: 6 June 2015
Published: 20 June 2015
Choice egg-count bioassays are a popular tool for analysing oviposition substrate preferences of gravid mosquitoes. This study aimed at improving the design of two-choice experiments for measuring oviposition substrates preferences of the malaria vector Anopheles gambiae senso lato, a mosquito that lays single eggs.
In order to achieve high egg-laying success of female An. gambiae sensu stricto (s.s.) and Anopheles arabiensis mosquitoes in experiments, four factors were evaluated: (1) the time provided for mating; (2) the impact of cage size, mosquito age and female body size on insemination; (3) the peak oviposition time; and, (4) the host sources of blood meal. Choice bioassays, with one mosquito released in each cage containing two oviposition cups both with the same oviposition substrate (100 ml water), were used to measure and adjust for egg-laying characteristics of the species. Based on these characteristics an improved design for the egg-count bioassay is proposed.
High oviposition rates [84%, 95% confidence interval (CI) 77–89%] were achieved when 300 male and 300 blood-fed female An. gambiae s.s. were held together in a cage for 4 days. The chances for oviposition dropped (odds ratio 0.30; 95% CI 0.14–0.66) when human host source of blood meal was substituted with a rabbit but egg numbers per female were not affected. The number of eggs laid by individual mosquitoes was overdispersed (median = 52, eggs, interquartile range 1–214) and the numbers of eggs laid differed widely between replicates, leading to a highly heterogeneous variance between groups and/or rounds of experiments. Moreover, one-third of mosquitoes laid eggs unequally in both cups with similar substrates giving the illusion of choice. Sample size estimations illustrate that it takes 165 individual mosquitoes to power bioassays sufficiently (power = 0.8, p = 0.05) to detect a 15% shift in comparative preferences of two treatments.
Two-choice egg count bioassays with Anopheles are best done with a two-tier design that (1) implements a parallel series of experiments with mosquitoes given a choice of two identical substrates choices and, (2) uses a single mosquito in each test cage rather than groups of mosquitoes to assess the preference of a test or control solution. This approach, with sufficient replication, lowers the risk detecting pseudopreferences.
Anopheles mosquitoes are efficient and resilient vectors of human malaria and filariasis in Africa. These vectors are mainly controlled by extensive use of long-lasting insecticidal nets (LLINs) and indoor residual spraying (IRS) of houses . The two interventions exploit the tendency of Anopheles funestus senso lato (s.l.) and Anopheles gambiae s.l., the major vectors of malaria in sub-Saharan Africa, to bite and rest indoors  and have together contributed to a remarkable and consistent decline in the transmission of malaria through the last decade [3–6]. However, like all vector control interventions, these too have limitations and when used in isolation could fall short in areas with: (1) strains selected for physiological resistance to insecticides [7–10]; (2) secondary vector species that live and bite outdoors [11, 12]; and, (3) cryptic vector sub-groups that bite in the early evening and/or bite outdoors . These and a complex of other factors, including increasing drug resistance and high costs of interventions, make malaria resurgence a grim reality . New strategies with novel tools that combine with LLINs and IRS to target these elusive groups of vectors in addition to the major vectors could prevent the resurgence of disease and hasten malaria elimination.
Larval source management (LSM) can be a complementary intervention for targeting all strains of malaria vectors irrespective of their state of insecticide resistance or resting and biting tendency. However in areas with extensive oviposition sites LSM becomes challenging [15, 16]. Attempts to target oviposition sites by identifying precisely the physical features of water bodies with mosquito larvae have so far been unsuccessful [17, 18]. Nevertheless field studies suggest that the presence of early instar larvae in water bodies is non-random, which may indicate that gravid females select particular water bodies in which to lay their eggs. These studies imply that favourable aquatic oviposition sites though highly heterogeneous in form, space and time [17–20] will display key features that act as signature cues for gravid mosquitoes seeking to lay eggs. Identifying the cues that elicit oviposition behaviour could aid the targeting of larvicides into productive mosquito oviposition sites and allow the development of odour-baited gravid mosquito traps for Anopheles.
Laboratory experiments within insect cages are a simple first step in identifying cues that guide short-range habitat selection in gravid mosquitoes . Of these experiments, choice egg-count bioassays are the most common and have been used to search for cues that are preferred or avoided by mosquitoes seeking to lay eggs [22, 23]. Here eggs or egg rafts laid in test substrates by groups of mosquitoes are counted and compared to those laid in a reference substrate, the control. Using these choice tests, chemicals that influence oviposition have been identified for Stegomyia [24–26], Culex [25, 27–29], and recently for An. gambiae sensu stricto (s.s.) [30, 31]. In addition egg-count bioassays have been used to investigate the response of gravid An. gambiae s.s. to bacterial cultures [32–34] with variable outcomes. All choice egg-count experiments with An. gambiae s.l. have been prepared, implemented and reported in a unique way making it difficult to generalize findings.
Already McCrae raised concern  that the very low mean numbers of eggs commonly reported in many choice egg-count experiments with groups of An. gambiae s.l. mosquitoes suggest that very few test mosquitoes actually lay eggs and that “the behaviour of only three or four mosquitoes was [therefore] tested”. This could lead to invalid conclusions, should it be true. In addition, it must be considered that putative oviposition substrates may consist of age-dependent organic infusions or concentration-sensitive chemicals, which degrade in time and should be evaluated over a short period. This study sought to improve on the preparation and establish the peak oviposition time of gravid mosquitoes to ensure that most laid eggs during experiments and were tested when they are likely to be most receptive to cues important for oviposition.
It was hypothesized that the egg-laying behaviour of An. gambiae s.l. makes it necessary to re-design choice egg-count bioassays uniquely for this species . The need to consider the design of these bioassays for species that lay single eggs and exhibit skip oviposition has been well shown by Chadee and Corbet  who proposed a new study design for Aedes: one that entailed “recording the distribution of eggs by individual females initially provided with an array of identical sites”. However, their work has been ignored in consequent studies with this genus [38–42]. The present study therefore aimed to present new approaches for: (1) optimizing egg laying in test insects; (2) implementing tests to compare two substrates; and, (3) analysing and reporting finding of egg-count experiments.
Mosquito rearing procedures
Insectary-reared An. gambiae s.s. and Anopheles arabiensis (Mbita strains) were used for this study. Briefly, 2–3 days old mosquitoes were allowed to feed on a human arm for 15 min on two consecutive evenings at 19:00 hours. On the third day, blood-fed females were allowed to oviposit on wet filter papers provided overnight in the cage. Eggs were dispensed in 20-L plastic tubs (41 cm diameter, 8 cm deep) half-filled with non-chlorinated tap water purified by filtering through a charcoal-sand filter (hereafter called tap water). Hatched larvae were fed with ground Tetramin® baby fish food (Tetra, Melle, Germany) twice daily. Pupae were collected into 10-cm diameter plastic cups filled with 200 ml of tap water and left overnight in mosquito cages for adults to emerge. Adults were maintained on 6% glucose ad libitum using absorbent paper wicks propped in 25-ml vials filled with glucose solution.
Females were immobilized by placing them in a refrigerator at 4°C for 15 min. Terminalia and the near terminal abdominal segment (segment IX) were severed in normal saline to expose spermathecae. Slide mounts of spermathecae were inspected using a microscope at 1,000× magnification for the presence of motile spermatozoa—a confirmation for insemination. The abdominal segments VII and VIII were gently severed to expose ovaries. The ovaries were observed at a magnification of 200× to evaluate stages of egg development. Mosquitoes with mature eggs, boat-shaped with fully developed floats, were categorized as gravid. To estimate the size of a female, the length of the left wing was measured from the axial incision to the tip (omitting the fringe setae) to the nearest 0.1 mm [43, 44].
Cages and oviposition cups
Experiments were carried out in standard cages (30 × 30 × 30 cm) or in large cages (60 × 60 × 60 cm). The cages had a steel framework on a galvanized metallic base and covered with fine cotton mosquito netting. The cage net also had an insert sleeve for introducing and retrieving oviposition substrates and gravid mosquitoes. Oviposition substrates were offered in 7-cm diameter, 100-ml clear borosilicate crystallising glasses (Pyrex®, hereafter called oviposition cups). Prior to any experiment oviposition cups were autoclaved and kept at 200°C for 2 h to reduce the possibility of bacteria and odorant contamination. Individual gravid mosquitoes (indicated by an enlarged, pale white abdomen) were introduced into the cages and provided with either one or two oviposition cups containing 100 ml of tap water.
Experimental procedures I: improving the egg laying success of Anopheles gambiae s.s. for egg-count experiments
It is important to produce large numbers of gravid mosquitoes that are consistent in their egg-laying behaviour during oviposition experiments. Here a series of experiments was carried out with caged local mosquito strains to evaluate and optimize major factors that are known to affect egg laying of mosquitoes in laboratory colonies including mating and insemination [45, 46], host source of blood meal [47–51] and the periods between blood feeding and the provision of suitable substrates [52–54].
Providing sufficient time for mating
Relatively young females are selected and blood fed when preparing gravid mosquitoes for experiments to maximize the proportion that survive until experiments are done 2–3 days after the blood meal. It was explored if young females, separated from males after 3 days, and then blood fed are sufficiently inseminated to develop and lay eggs 3 days later and whether the oviposition success can be improved by holding females with males after the blood meal. Two groups of 300 3-day old female An. gambiae s.s. mosquitoes were put in separate standard cages. These mosquitoes had spent the first 3 days of their adult life in colony cages with over 1,000 male and female conspecific mosquitoes. In one cage 300 sibling males of the same age were added, whilst in the other cage no male mosquitoes were included. Both groups of mosquitoes were then starved of sugar solution for up to 6 h. Tap water saturated cotton towels folded to a size of 50 × 25 cm were placed over the cages to maintain the relative humidity (RH) between 68 and 75%. The starved mosquitoes were permitted to blood feed from a human arm at twilight (between 18:30 and 19:30) for 15 min in imposed total darkness. Mosquitoes that were not fed after the first blood meal were removed from the cage, killed and discarded. Sugar solution was then replaced in the preparation cage until 12:00 the following day when the mosquitoes were starved again in preparation for a second blood meal the same evening. After blood feeding the mosquitoes were left in the insectary at temperatures that averaged 27 ± 2°C and were only retrieved 72 h later at the onset of experiments. The mosquitoes were 7 days old when egg-count cage experiments were implemented. Fifty females were selected based on their abdominal appearance from each of the two cages and transferred individually to standard cages. Each female was offered a single oviposition cup with 100 ml of tap water. The presence and number of eggs was recorded after 16 h (17:00–08:00). This experiment was carried out on three occasions (rounds; 3 × 50 individual females per treatment). Identical experiments were carried out with An. arabiensis.
Impact of cage size, mosquito age and body size on insemination
The previous experiment revealed a very low (<30%) oviposition rate in An. arabiensis even when kept with males for 7 days. Therefore, an experiment was designed to investigate the role of cage size, age and size of mosquitoes on the insemination success (proportion of inseminated females) of An. gambiae s.s. and An. arabiensis. Experiments were carried out in standard and large cages in parallel. For both species, 1,400 pupae were placed in plastic cups (10 cm diameter) filled with 200 ml of tap water positioned in each of two standard cages for 24 h. From that cage, 300 newly emerged male and 300 female mosquitoes of each species were transferred into separate cages of the two sizes. Six per cent glucose solution was provided in all cages ad libitum. After 3 days, 25 female mosquitoes were randomly selected from each cage by a technician unaware of the objectives of the study and dissected to evaluate insemination and to measure the wing lengths. The same number of females was dissected for days 4, 5 and 6. The experiment was implemented for three rounds with different batches of mosquitoes. Mosquitoes in this experiment were not offered blood meals.
Optimal timing of cage experiments
To have a consistently large proportion of females respond in oviposition experiments, it was important to establish the optimum interval between the last blood meal and the bioassay for the local mosquito strains. Furthermore, putative test substrates for oviposition in mosquitoes (e.g., bacteria solutions, plant infusions, volatile inorganic compounds) are often unstable. It is therefore important to target the experiments just before the peak in egg laying. Cage experiments were carried out with different mosquitoes 48 and 72 h after their second blood meal. ‘Gravid’ mosquitoes for experiments were prepared following standard procedures. For each experiment, 100 An. gambiae s.s. mosquitoes were individually offered two oviposition cups with tap water in two-choice egg-count experiments at 17:00. In 50 of the 100 cages, both oviposition cups were retrieved and replaced with two new cups containing tap water at 21:30, the remaining 50 cages were left undisturbed through the night. The aim here was to investigate if the caged An. gambiae s.s. have several oviposition peaks during the night and to explore when skip oviposition occurs. Half the cages remained undisturbed as a control to investigate if the exchange of cups might interfere with the oviposition response during the night. The experiment was ended at 08:00 the following morning and the number of eggs in each cup recorded. Both experiments were carried out for three rounds.
Understanding the impact of host sources of blood meal on egg laying
Anopheles gambiae s.s. is highly anthropophagic  and there is evidence that different host sources of blood meals have an impact on the oviposition rate and fecundity  of this species. An experiment was designed to elucidate the impact of feeding caged An. gambiae s.s. on non-human hosts on the proportion of females becoming gravid and the number of eggs laid by each female. Different groups of mosquitoes were blood-fed on either a human arm or rabbit. Blood meals on human arm were offered as described in the previous experiment. For rabbit host blood meals, fur was shaved on the ventral side of the rabbit in an area of 15 × 5 cm (approximately equal to the area exposed by an extended human arm covered with a latex glove). The rabbit was then held in a restrainer that limited movement and exposed the shaven underside. Mosquitoes to be fed were held in a cage positioned at the base of the restrainer allowing free access to the shaven area. In each treatment a group of 300 females were fed on two consecutive days. All mosquitoes that did not blood feed on the first day were removed from the cages. The blood-fed mosquitoes were then held together with 300 males in standard cages for 72 h. A total of 100 female mosquitoes were randomly selected from each of the two cages, aspirated out by a technician unaware of the objectives of the research, and dissected to determine if they were gravid or not. Another 25 females visually appearing gravid were purposively selected by an experienced technician from each of the two cages. These 25 females were tested individually in no-choice egg-count cage experiments to compare the proportion of females that laid eggs and the number of eggs laid per female fed on either rabbit or human blood. The experiment was done for three rounds using different batches of mosquitoes.
Experimental procedures II: improving the experimental design of cage egg-count bioassays with Anopheles gambiae s.s.
To promote an empirical evaluation of substrate preferences using two-choice egg-count bioassays it is important to understand the natural egg-laying pattern of An. gambiae s.s. and take it into account when designing experiments. This experiment was aimed at identifying an appropriate design for egg-count oviposition studies with An. gambiae s.s. and at highlighting the importance of sample size. Specifically (1) the number of eggs laid by individual An. gambiae s.s. mosquitoes was estimated and their statistical distributions and variances explored; (2) skip-oviposition within experimental cages was quantified; and, (3) the variability in egg counts and response towards two equal choices of oviposition substrate analysed between rounds.
Two equal choice egg-count bioassays with individual gravid females to explore egg distribution and variability in egg-counts
Sample size considerations
When implementing two-choice bioassays with two different oviposition substrates the assumption is that there is no preference between the two substrates. However, it is likely that the chances of a type 1 error (i.e., artefact preference for one substrate over then other) are increased with a small number of replicates. Therefore the data were used to estimate the sample size required for routine bioassays using power calculation for two-sample comparisons of proportions (power.prop.test function in R software) and for a single proportion compared to a known proportion . A 50% distribution was assumed when equal choices are presented (p1 = 0.5). Power estimates were generated to predict a 15% increase or decrease (p2 = 0.65) in oviposition response to a test medium at sample sizes between five and 225 and generated estimates of effect sizes generated that can be detected with 80% power at a 5% significance level for the same range of sample sizes.
Multivariable analyses were implemented using generalized estimation equations (GEE) to analyse how proportions were affected by test variables. Different batches of mosquitoes from different rounds of an experiment were considered clustered (not independent) and included in the GEE model as repeated measure. To evaluate the impact of including male mosquitoes in cages with blood-fed females on the proportions of mosquitoes that lay eggs, a GEE model was fitted with binomial distribution, logit link function and exchangeable correlation matrix. The presence of male mosquitoes (coded a 1 if a male is present, and 0 otherwise) in the holding cages was included in the model as a fixed factor. The association between the proportion of female mosquitoes inseminated and cage size (standard = 0, large = 1), mosquito age (a four-level categorical variable coded 3, 4, 5, and 6 days) and mosquito size (measured in terms of wing length) was assessed using two separate GEE models (binomial distribution, logit link function, exchangeable correlation matrix) for the two test species. An interaction term of age and cage size was also included in the model. Similar models were fitted for the experiment on host blood meal sources with the blood meal source as a fixed factor, and the experiment on oviposition time with time period as a fixed factor. The mean numbers of eggs and their corresponding 95% confidence intervals (CIs) for two treatments were calculated as the exponent of the parameter estimates based on generalized linear models with negative binomial distributions with no intercept included.
Egg numbers laid by individual females were tested for normality using the Kolmogorov–Smirnov test. Additionally, overdispersion (i.e., meaning the variability in the data is not equal to the mean, as in the Poisson distribution) was assessed by inspecting the residual deviance, which follows a Chi-squared distribution where the expected value should be close to the degrees of freedom if the data are not overdispersed. Overdispersion is a problem because it may cause standard errors of the estimates to be deflated or underestimated, i.e., a variable may appear to be significant when in fact it is not. The assumption of homogeneity of variance in the correlated count data collected from control and test cups was tested with the Pitman–Morgan test .
In the two equal choice egg-count bioassay one mosquito were presented with two cups with tap water for oviposition. The data derived from these two cups per individual mosquito were related and therefore a mosquito was considered a cluster in the GEE models. GEE models assuming exchangeable working correlation and with a negative binomial distribution with a log link function fitted were used to explore differences in egg counts between control and test cups and between rounds (fixed factors), whilst GEE models with a binomial distribution and logit link faction fitted were used to estimate the odds of a female choosing the test cup over the control. All mean counts or mean proportions per treatment and their 95% CIs were calculated as the exponential of the parameter estimates for models with no intercept included.
Ethical approval for this study was obtained from the Kenya Medical Research Institute’s Ethical Review Committee (Protocol no. 422).
Including males in holding cages after blood meals increases the proportion of ovipositing females
The odds of a female An. gambiae s.s. laying eggs were nine times greater if, after a blood meal, she was held with males than without them (OR = 9.0, 95% CI 7.9–9.5, p < 0.01). On average 84% (95% CI 77–89%) of females laid eggs when held with males compared to 36% (95% CI 29–44%) when held without. Whilst the total number of eggs laid by females held with males [2,904 eggs (95% CI 2,844–2,968)] was three times as high as the total number laid by females kept separated from males after blood-meals [994 eggs (95% CI 959–1,030)], the mean number of eggs laid per female was similar in mixed-sex cages [66 eggs, 95 % CI 44–99 eggs] and female only cages (54 eggs, 95% CI 36–82) highlighting the benefit of observing individual rather than groups of mosquitoes.
Only 25% of An. arabiensis (95% CI 15–41%) laid eggs. The likelihood of laying eggs was not associated with the presence or absence of males in the cages after blood feeding (OR 1.92, 95% CI 0.62–5.98, p = 0.658) and dissections showed that a large proportion (>50%) of females that did not lay eggs were not inseminated. The mean number of eggs laid per female that laid was 63 (95% CI 59–68).
Age and body size can impact on insemination success in Anopheles gambiae and Anopheles arabiensis irrespective of cage size
Multivariable analysis of factors tested in association with the rate of insemination
Anopheles gambiae s.s.
OR (95% CI)
OR (95% CI)
Age of mosquito in days
Interaction between mosquito age and cage size
Egg-count cage bioassays are best implemented 72 h after the last blood meal during the peak oviposition time between 17:00 and 21:30
Evaluation of egg laying periodicity in caged Anopheles gambiae s.s. (Mbita strain)
Percentage of mosquitoes that laid eggs (95% CI)
Duration after blood meal (cups left overnight) (h)
Egg laying period (72 h after blood meal)
Skip oviposition (72 h after blood meal)
Blood meal sources from a secondary host reduces the proportion of females that become gravid but does not affect egg numbers laid by gravid females
Effect of host source of blood meal on oviposition of caged Anopheles gambiae s.s. (Mbita strain)
Mean (95% CI)
OR (95% CI)
Percentage of blood-fed mosquitoes gravid at dissection
Percentage of gravid mosquitoes that laid eggs in cage experiments
Number of eggs per gravid female that laid
Individual Anopheles gambiae s.s. lay a highly variable number of eggs despite standardized preparation procedures
The response of gravid females presented with two equal choices can be skewed when egg-counts are compared
In total 77,664 eggs were laid by 1,443 mosquitoes tested individually over 41 rounds; 41,113 (53%) eggs were laid in cups randomly labelled as test cups, and 36,551 (47%) in control cups. In addition to the overdispersed distribution of eggs, the two correlated variances of the egg counts in control and test cups were not homogeneous (p < 0.01). Generalized linear modelling with a negative binomial distribution fitted indicated that the differences in egg counts between control and test cups were small but statistically significant [rate ratio (RR) 1.13 (95% CI 1.01–1.25, p < 0.01)]. Furthermore, counts differed significantly between rounds (p < 0.001) with mean number of eggs laid per female in a cup in different rounds ranging between 17 (95% CI 13–20) and 46 (95% CI 39–55).
Comparing the proportional distribution of eggs leads to more reliable inference than using absolute egg counts
High between-cage variability in egg-counts and proportions must be expected when testing small groups of gravid females in egg-count cage bioassays
One-third of gravid Anopheles gambiae s.s. distribute their eggs in more than one oviposition medium (‘skip-oviposition’)
To detect an increase in oviposition response of 15% as compared to the baseline proportion (80% power and 5% significance) at least 165 responders need to be tested in each treatment group
Based on sample size calculations for two independent proportions, 165 responders need to be tested in each group (165 for p1 and 165 for p2; total 340) to detect an increase or decrease in oviposition response of 15% (p2 = 0.65) compared to the baseline proportion (p1 = 0.50) at 80% power and 5% significance. With a smaller sample size the effect size that can be detected increases, i.e., 90 replicates in each treatment arm can detect a difference between the proportions of not less than 20% (p1 = 0.50 and p2 = 0.70) and 30 replicates of not less than 33% (p1 = 0.50 and p2 = 0.83) (Figure 10).
These sample size considerations apply irrespective of whether the proportions of eggs laid by groups of mosquitoes per cage or by individual mosquitoes per cage are compared since in both cases only a single data point per cage can be recorded and the proportion of non-responders in the cage is unknown. Nevertheless, if large groups (>30 responders per cage for example) are used where the baseline proportion can be predicted to be close to 50% with some certainty it might be justifiable to use the sample size calculation for the inference for a single proportion comparing to a known proportion (0.5). In this case 85 replicate cages would be required for detecting a 15% increase compared to the baseline proportion at 80% power and 5% significance (Figure 10). Whilst this number of replication appears to be considerably lower it needs to be observed that over seven times more gravid females would be required in this experimental design (85 × 30 = 2,550) than when using individual females and two treatment arms (2 × 165 = 330).
Improving the egg-laying success of gravid females for egg-count experiments
Consistent with previous studies, insemination was shown to be important for egg laying by An. gambiae s.l. [45, 46]. The proportion of test females that laid eggs in the bioassays more than doubled when they were held in cages with males after blood feeding providing a longer period to mate. This gives further evidence that at least in laboratory settings mating in this species continues after the females have taken a blood meal. Depending on the age of insects at blood feeding this could be of great consequence. Cages with blood-fed females must be conditioned for insemination by including male mosquitoes especially when test females are blood fed at a relatively young age (here 2–3 days). In experiments evaluating the rate of insemination with non-blood-fed mosquitoes it was shown that approximately one-fifth of An. gambiae s.s. were still virgins when 6 days old (the average age of test mosquitoes across studies). This might explain the similar proportion of test mosquitoes that failed to lay eggs in the bioassays even under optimized preparation procedures. Increasing the number of males in cages might improve both the rates of insemination and egg laying . However, Verhoek and Takken  have demonstrated that ratios of 3:1 male to female do not significantly improve the rate of mating over a 1:1 ratio for An. gambiae s.l.
Rabbit blood meals resulted in a lower proportion of mosquitoes that eventually became gravid. This suggests that the common practice of substituting human hosts with rabbits, and possibly other secondary host sources of blood [63, 66, 67] potentially reduces the number of gravid mosquitoes and therefore increases the risk of including mosquitoes that will not lay eggs in bioassays. Excluding mosquitoes that did not lay eggs from the analysis showed that the actual mean number of eggs laid per female that became gravid after the blood meal was the same irrespective of the source of blood. If groups had been tested instead a false lower mean numbers of eggs with rabbit blood meals would have been inferred. Great caution is advised in selecting gravid mosquitoes where secondary host sources of blood are used in preparing test mosquitoes. By using individuals it is possible to implement choice test even where the impact of the host-source of blood meal is large or unknown. Mosquitoes that do not lay eggs can be removed from the final data set and reported as a separate entity of interest.
McCrae  wrote “For any study of oviposition to be complete it would be valuable to know the probable time of its occurrence as a basic guide for laboratory procedures”. The vast majority of the Mbita strain of An. gambiae s.s. did not yet lay eggs 48 h after the last blood meals; egg laying was constrained to early evening hours of the third night (≈72 h) after blood meals. This confirmed the findings of Haddow and others . Consequently, egg-count cage bioassays with the Mbita strain are best done between 17:00 and 21:30 on the third night after the last blood meal. However, controversial results have been published in the past. Other studies with An. gambiae s.s. have shown that some strains are laying eggs 48 h after a blood meal and it was suggested that egg-laying times depend on local conditions, blood-feeding times and temperature . Some studies also show that An. gambiae s.s. can lay eggs at any time throughout the dark phase [53, 54]. In consideration of these divergent findings, it is strongly recommended that oviposition periodicity studies precede all oviposition studies with different strains of this species. This does not only apply to egg-count cage bioassays but is equally important when investigating chemoreception in gravid females and changing sensilla sensitivity in response to changes to the physiological stage of a mosquito. These studies are often done 24 and 48 h after a blood meal [30, 69], which might not necessarily coincide with the time a female searches for an oviposition site.
The insectary-reared Mbita strain of An. arabiensis showed low rates of insemination compared to An. gambiae s.s. from the same area. At best 45% of all female An. arabiensis mosquitoes were inseminated after 6 days when held with an equal number of males throughout. There is some evidence that An. arabiensis is more difficult to mate and colonize in the laboratory compared to An. gambiae s.s. , although others have shown contrasting results where the rate of insemination in An. arabiensis of every age between 1 and 7 days was higher than that of An. gambiae s.s. . Their findings were probably due to longer colonization of the strain which selected for this trait. Increasing the size of holding cages to increase mating activity and insemination success in An. arabiensis was without significant gain. Low insemination and consequently low oviposition rates make it difficult to study the oviposition response of An. arabiensis to different oviposition substrates. Especially, when groups of An. arabiensis are used, caution should be exercised in interpreting the results by examining the mean egg numbers critically to ensure that the majority of the exposed females actually laid eggs. It has been shown that larger females were more likely to be inseminated compared to smaller ones. Attempting to optimize larval rearing conditions to increase adult body size and selecting for the largest females from the colony cages for experiments might thus be a reasonable approach to increasing oviposition rates in egg-count cage bioassays.
Improving the experimental design of cage egg-count bioassays with Anopheles gambiae s.s.
Using two equal choice egg-count bioassays with individual gravid mosquitoes illustrated the importance of (1) suitable experimental design based on the behavioural ecology of An. gambiae s.s.; (2) estimated sample sizes; and, (3) appropriate statistical analyses (Figure 11). This study confirmed that egg counts of individual female An. gambiae s.s. of the same age fed on the same source of blood and reared under standardized conditions are highly variable and overdispersed. Lyimo and Takken  previously demonstrated that individual newly emerged An. gambiae s.l. of the Muheza strain laid between 48 and 178 (mean 111) eggs while wild field populations laid an equally variable 66–290 (average 150) eggs. Hogg and Hurd  later confirmed variations in egg numbers showing that wild An. gambiae s.s. and An. arabiensis of Gambia laid between 20 and 180 eggs and five and 160 eggs, respectively. These wide disparities in egg numbers of individual females have also been shown for laboratory strains of other Anophelinae including: Anopheles stephensi , Anopheles sergenti , Anopheles multicolor, and Anopheles pharoensis . Suleman and others  noted that a small portion of An. stephensi females laid a very high number of eggs per batch, leading to a negative binomial distribution as also well demonstrated for An. gambiae s.s. in this study. Similar heterogeneity in egg numbers between individual females have also been shown for Aedes aegypti . This may be a general trait of mosquitoes that lay single eggs, rendering the use of egg numbers to gauge oviposition substrate preferences inappropriate especially with small groups of mosquitoes . It was demonstrated that the high variation in the number of eggs laid by individual females can lead to an unequal distribution of eggs in equal substrates. This disproportion persisted even with very large sample size.
Exploring the pattern of ‘skip oviposition’ in An. gambiae s.s., it was demonstrated that approximately one-third of all gravid An. gambiae s.s. distribute their eggs in more than one oviposition site, a behaviour that is well known in Aedes mosquitoes [77, 78], but has been poorly described in An. gambiae s.l. species [36, 79] in laboratory egg-count experiments, possibly because most experimenters use groups of mosquitoes, which masks skip oviposition. There is also indirect evidence of skip oviposition from one study in the field  showing that this is not an artefact trait of colonized mosquitoes but rather an inherent trait of the species. Skip oviposition represents a response of the gravid female to the substrates and should not be excluded from analyses. Skip ovipositing females choose to use both substrates, therefore not rejecting any, an important event with reference to comparative preference of substrates. Importantly, An. gambiae s.s. females do not distribute their eggs in equal proportions but in most cases lay two-thirds in one and one-third in the other oviposition cup. Since observations in this study are based on equal choices, it is clear that the higher egg batch does not indicate a preference. It is important to note that individual skip ovipositing female did not lay more eggs compared to those individual females that laid in a single cup.
In experiments, where groups of females are analysed in oviposition assays, the marked heterogeneity of egg numbers laid by individual females combined with skip oviposition is likely to increase the variance in the system and this could lead to a type 1 error where an unequal distribution of eggs between the test and control solutions is wrongly considered to be true, especially if group sizes are small. Here it was illustrated that this frequently happens when group numbers per cage are below 30 responders. Considering that of those, probably a fifth or more mosquitoes do not lay eggs, a skewed distribution can be expected and only a large number of cages can be able to detect true differences of substrates. Since many choice experiments with anophelines are done with groups much lower than 30 [22, 30, 35, 61], results need to be interpreted with caution.
This study demonstrated that observing individual mosquito’s responses to oviposition substrates rather than groups has a number of advantages. This approach ensures that only responders are included in the data analysis. It allows the analysis of choice based on a binary outcome, the enumeration of egg numbers of individual females and the observation of skip oviposition, which has previously been shown to be influenced by the suitability of a substrate . Last but not least, the necessary number of replications can be achieved with a smaller number of gravid females compared to when groups are used.
Sample size considerations are rarely reported for entomological studies and the number of replications hardly ever justified in publications. This study illustrates that insufficient replication might not only hamper the ability to show a significant effect due to the lack of power, but also demonstrates that a small number of replicates and small group sizes can result in significant artefact differences in oviposition responses in two-choice experiments purely based on stochastic effects rather than due to a treatment effect. Misinterpretation of results can be reduced by sufficient replication and validation of the experiment by implementing a control experiment preferably in parallel .
The underlying hypothesis of a choice experiment is that when two (or more) equal choices are presented the response towards these choices is equally proportional with odds of success of 1:1 (baseline or control). Choices by virtue of the design of the experiment should be analysed as proportions rather than absolute counts, especially when count data are highly variable. If an oviposition cue is presented that is either preferred or avoided by gravid females a significant diversion from the baseline is expected. It was shown that there is a high variability in the response towards a test and control cup containing the same substrate in individual rounds of experiments highlighting the importance of large sample sizes and the implementation of an experiment over several rounds with different batches of mosquitoes. The behaviour of mosquitoes from the same batch might be affected for example by their rearing history and/or by the climatic conditions during the experiment or other non-measurable random effects. Replicate tests with mosquitoes from the same batch implemented on the same day with the same batch of oviposition substrate should not be considered independent; it is pseudo-replication . In order to document the baseline, including its 95% CI, it is recommended that choice experiments with different test substrates in a cage must always be implemented in parallel with a control experiment with the same number of equal choices. This validates the experimental design  and allows statistical comparison of the odds of success in the test experiment with the odds of success in the control experiment (baseline).
The classic oviposition index represents a proportional comparison of the numbers of eggs, egg rafts or females  but is rarely used in oviposition experiments with An. gambiae s.s.. Frequently the mean number of eggs in test and control cups is compared using classical ANOVA and t tests [30, 34, 35, 61, 63, 66, 67, 83–87]. These assume normality of data distribution and homogeneity of variance , but both assumptions are violated when looking at egg counts of An. gambiae s.s.. Some (log-) transform egg-counts or use non-parametric tests that do not assume a normal distribution. However, log-transforming count data for analyses have recently been challenged except when dispersion is small and means are large . Moreover, non-parametric have reportedly been invalidated even by “small differences in variance and moderate degrees of skew” [90–92]. When distributions are skewed (such as for negative binomial distributions) differences in means are prone to go together with differences in variance . It is also imperative to appreciate the non-independent nature of the data from control and test cups in the same cage and the dependent nature of the data derived from the same rounds when analysing choice egg-count bioassays. This violation of independent observations assumption results on downwardly biased standard error estimates, overly large test statistics, and inflated type I error rates. The statistical procedure used must, therefore, take account of that by including repeated measure terms.
It is strongly suggested analysing choice bioassays using generalized regression models that allow for the appropriate distribution to be fit to the model rather than transforming the data [88, 89]. Preference should be given to analysing proportions (of eggs laid or of females laying in test and control) using a binomial distribution than to analyse counts using a negative binomial or Poisson distribution. Importantly, these models allow including critical explanatory variables as well as random factors and/or repeated measures that might have affected the outcome. Based on the model, the effect size of the test can be reported using both odds ratios and predicted averages together with associated CIs .
Individual An. gambiae s.l. lay a widely different number of eggs. A large proportion of these also skip oviposit, spreading their eggs in more than one substrate. These egg-laying patterns can lead to spurious conclusions of oviposition substrate preferences based on two choice egg-count bioassays. In order to increase the accuracy of these bioassays, designs that take into account the natural variability in the number of eggs and ensure sufficient replication are needed. Experiments are most accurate when gravid females are prepared and selected under carefully controlled conditions and when implemented in a two-tier design with 165 individual mosquitoes in each treatment arm: 165 cages each with one mosquito given a choice between a test and control solution and 165 similar cages where the mosquito has a choice between two identical control solutions. This will enable the detection of differences in substrate preferences of as little as 15% with sufficient statistical power and significance.
MNO, UF, JML, SWL, and ST conceived the idea for this research and jointly developed the experimental designs. MNO developed all protocols and implemented the experiments together with EM. MNO and UF analysed the data and drafted the manuscript with support from BO. SWL, JML, BO, and ST contributed to the final draft. All authors read and approved the final manuscript.
We thank Benard Oyembe, Elisha Obudho and David Alila for providing mosquitoes for this study and Paul Ouma, Rose Atieno and Gregory Masinde for their technical support. We also extend our gratitude to Stephen Young for statistical advice. This project was funded by the National Institute of Health (NIH) through grant no. R01AI082537.
Compliance with ethical guidelines
Competing interests The authors declare that they have no competing interests.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Enayati A, Hemingway J (2010) Malaria management: past, present, and future. Annu Rev Entomol 55:569–591PubMedView ArticleGoogle Scholar
- Takken W, Knols BG (1999) Odor-mediated behavior of Afrotropical malaria mosquitoes. Annu Rev Entomol 44:131–157PubMedView ArticleGoogle Scholar
- Nyarango PM, Gebremeskel T, Mebrahtu G, Mufunda J, Abdulmumini U, Ogbamariam A et al (2006) A steep decline of malaria morbidity and mortality trends in Eritrea between 2000 and 2004: the effect of combination of control methods. Malar J 5:33PubMed CentralPubMedView ArticleGoogle Scholar
- Ceesay SJ, Casals-Pascual C, Nwakanma DC, Walther M, Gomez-Escobar N, Fulford AJ et al (2010) Continued decline of malaria in The Gambia with implications for elimination. PLoS One 5:e12242PubMed CentralPubMedView ArticleGoogle Scholar
- O’Meara WP, Mangeni JN, Steketee R, Greenwood B (2010) Changes in the burden of malaria in sub-Saharan Africa. Lancet Infect Dis 10:545–555PubMedView ArticleGoogle Scholar
- Mharakurwa S, Mutambu SL, Mberikunashe J, Thuma PE, Moss WJ, Mason PR et al (2013) Changes in the burden of malaria following scale up of malaria control interventions in Mutasa District, Zimbabwe. Malar J 12:223PubMed CentralPubMedView ArticleGoogle Scholar
- Chandre F, Darrier F, Manga L, Akogbeto M, Faye O, Mouchet J et al (1999) Status of pyrethroid resistance in Anopheles gambiae sensu lato. Bull World Health Organ 77:230–234PubMed CentralPubMedGoogle Scholar
- Chandre F, Darriet F, Manguin S, Brengues C, Carnevale P, Guillet P (1999) Pyrethroid cross resistance spectrum among populations of Anopheles gambiae s.s. from Cote d’Ivoire. J Am Mosq Control Assoc 15:53–59PubMedGoogle Scholar
- Ranson H, Abdallah H, Badolo A, Guelbeogo WM, Kerah-Hinzoumbe C, Yangalbe-Kalnone E et al (2009) Insecticide resistance in Anopheles gambiae: data from the first year of a multi-country study highlight the extent of the problem. Malar J 8:299PubMed CentralPubMedView ArticleGoogle Scholar
- Ranson H, N’Guessan R, Lines J, Moiroux N, Nkuni Z, Corbel V (2011) Pyrethroid resistance in African anopheline mosquitoes: what are the implications for malaria control? Trends Parasitol 27:91–98PubMedView ArticleGoogle Scholar
- Russell TL, Govella NJ, Azizi S, Drakeley CJ, Kachur SP, Killeen GF (2011) Increased proportions of outdoor feeding among residual malaria vector populations following increased use of insecticide-treated nets in rural Tanzania. Malar J 10:80PubMed CentralPubMedView ArticleGoogle Scholar
- Tirados I, Costantini C, Gibson G, Torr SJ (2006) Blood-feeding behaviour of the malarial mosquito Anopheles arabiensis: implications for vector control. Med Vet Entomol 20:425–437PubMedView ArticleGoogle Scholar
- Riehle MM, Guelbeogo WM, Gneme A, Eiglmeier K, Holm I, Bischoff E et al (2011) A cryptic subgroup of Anopheles gambiae is highly susceptible to human malaria parasites. Science 331:596–598PubMed CentralPubMedView ArticleGoogle Scholar
- Cohen J, Smith D, Cotter C, Ward A, Yamey G, Sabot O et al (2012) Malaria resurgence: a systematic review and assessment of its causes. Malar J 11:122PubMed CentralPubMedView ArticleGoogle Scholar
- Fillinger U, Kannady K, William G, Vanek MJ, Dongus S, Nyika D et al (2008) A tool box for operational mosquito larval control: preliminary results and early lessons from the Urban Malaria Control Programme in Dar es Salaam, Tanzania. Malar J 7:20PubMed CentralPubMedView ArticleGoogle Scholar
- Fillinger U, Lindsay SW (2011) Larval source management for malaria control in Africa: myths and reality. Malar J 10:353PubMed CentralPubMedView ArticleGoogle Scholar
- Fillinger U, Sombroek H, Majambere S, van Loon E, Takken W, Lindsay SW (2009) Identifying the most productive breeding sites for malaria mosquitoes in The Gambia. Malar J 8:62PubMed CentralPubMedView ArticleGoogle Scholar
- Majambere S, Fillinger U, Sayer DR, Green C, Lindsay SW (2008) Spatial distribution of mosquito larvae and the potential for targeted larval control in The Gambia. Am J Trop Med Hyg 79:19–27PubMedGoogle Scholar
- Fillinger U, Sonye G, Killeen GF, Knols BG, Becker N (2004) The practical importance of permanent and semipermanent habitats for controlling aquatic stages of Anopheles gambiae sensu lato mosquitoes: operational observations from a rural town in western Kenya. Trop Med Int Health 9:1274–1289PubMedView ArticleGoogle Scholar
- Ndenga BA, Simbauni JA, Mbugi JP, Githeko AK, Fillinger U (2011) Productivity of malaria vectors from different habitat types in the western Kenya highlands. PLoS One 6:e19473PubMed CentralPubMedView ArticleGoogle Scholar
- Isoe J, Millar JG, Beehler JW (1995) Bioassays for Culex (Diptera: Culicidae) mosquito oviposition attractants and stimulants. J Med Entomol 32:475–483PubMedView ArticleGoogle Scholar
- Sumba LA, Guda TO, Deng AL, Hassanali A, Beier JC, Knols BGJ (2004) Mediation of oviposition site selection in the African malaria mosquito Anopheles gambiae (Diptera: Culicidae) by semiochemicals of microbial origin. Int J Trop Insect Sci 24:260–265View ArticleGoogle Scholar
- Bukhari T, Knols BGJ (2009) Efficacy of Aquatain™, a Monomolecular Surface Film, against the Malaria Vectors Anopheles stephensi and An. gambiae s.s. in the Laboratory. Am J Trop Med Hyg 80:758–763PubMedGoogle Scholar
- Bentley MD, McDaniel IN, Yatagai M, Lee HP, Maynard R (1979) p-Cresol: an Oviposition Attractant of Aedes triseriatus. Environ Entomol 8:206–209View ArticleGoogle Scholar
- Hwang YS, Schultz GW, Axelrod H, Kramer WL, Mulla MS (1982) Ovipositional repellency of fatty acids and their derivatives against Culex and Aedes mosquitoes. Environ Entomol 11:223–226View ArticleGoogle Scholar
- Sharma KR, Seenivasagan T, Rao AN, Ganesan K, Agarwal OP, Malhotra RC et al (2008) Oviposition responses of Aedes aegypti and Aedes albopictus to certain fatty acid esters. Parasitol Res 103:1065–1073PubMedView ArticleGoogle Scholar
- Hwang YS, Kramer WL, Mulla MS (1980) Oviposition attractants and repellents of mosquitoes: isolation and identification of oviposition repellents for Culex mosquitoes. J Chem Ecol 6:71–80View ArticleGoogle Scholar
- Millar JG, Chaney JD, Mulla MS (1992) Identification of oviposition attractants for Culex quinquefasciatus from fermented Bermuda grass infusions. J Am Mosq Control Assoc 8:11–17PubMedGoogle Scholar
- Millar JG, Chaney JD, Beehler JW, Mulla MS (1994) Interaction of the Culex quinquefasciatus egg raft pheromone with a natural chemical associated with oviposition sites. J Am Mosq Control Assoc 10:374–379PubMedGoogle Scholar
- Rinker DC, Pitts RJ, Zhou X, Suh E, Rokas A, Zwiebel LJ (2013) Blood meal-induced changes to antennal transcriptome profiles reveal shifts in odor sensitivities in Anopheles gambiae. Proc Natl Acad Sci USA 110:8260–8265PubMed CentralPubMedView ArticleGoogle Scholar
- Lindh JM, Okal MN, Herrera-Varela M, Borg-Karlson AK, Torto B, Lindsay SW et al (2015) Discovery of an oviposition attractant for gravid malaria vectors of the Anopheles gambiae species complex. Malar J 14:119PubMed CentralPubMedView ArticleGoogle Scholar
- Lindh JM, Kannaste A, Knols BG, Faye I, Borg-Karlson AK (2008) Oviposition responses of Anopheles gambiae s.s. (Diptera: Culicidae) and identification of volatiles from bacteria-containing solutions. J Med Entomol 45:1039–1049PubMedView ArticleGoogle Scholar
- Blackwell A, Johnson SN (2000) Electrophysiological investigation of larval water and potential oviposition chemo-attractants for Anopheles gambiae s.s. Ann Trop Med Parasitol 94:389–398PubMedGoogle Scholar
- Huang J, Miller JR, Chen SC, Vulule JM, Walker ED (2006) Anopheles gambiae (Diptera: Culicidae) oviposition in response to agarose media and cultured bacterial volatiles. J Med Entomol 43:498–504PubMedView ArticleGoogle Scholar
- McCrae AWR (1984) Oviposition by African malaria vector mosquitoes II. Effects of site tone, water type and conspecific immatures on target selection by freshwater Anopheles gambiae Giles, sensu lato. Trop Med 73:307–318Google Scholar
- Herrera-Varela M, Lindh J, Lindsay SW, Fillinger U (2014) Habitat discrimination by gravid Anopheles gambiae sensu lato—a push-pull system. Malar J 13:133PubMed CentralPubMedView ArticleGoogle Scholar
- Corbet PS, Chadee DD (1993) An improved method for detecting substrate preferences shown by mosquitoes that exhibit ‘skip oviposition’. Physiol Entomol 18:114–118View ArticleGoogle Scholar
- Bandyopadhyay P, Guha L, Seenivasagan T, Sathe M, Sharma P, Parashar BD et al (2011) Synthesis and bio-evaluation of aryl hydrazono esters for oviposition responses in Aedes albopictus. Bioorg Med Chem Lett 21:794–797PubMedView ArticleGoogle Scholar
- Ponnusamy L, Xu N, Boroczky K, Wesson DM, Abu Ayyash L, Schal C et al (2010) Oviposition responses of the mosquitoes Aedes aegypti and Aedes albopictus to experimental plant infusions in laboratory bioassays. J Chem Ecol 36:709–719PubMedView ArticleGoogle Scholar
- Ponnusamy L, Xu N, Nojima S, Wesson DM, Schal C, Apperson CS (2008) Identification of bacteria and bacteria-associated chemical cues that mediate oviposition site preferences by Aedes aegypti. Proc Natl Acad Sci USA 105:9262–9267PubMed CentralPubMedView ArticleGoogle Scholar
- Seenivasagan T, Sharma KR, Ganesan K, Prakash S (2010) Electrophysiological, flight orientation and oviposition responses of three species of mosquito vectors to hexadecyl pentanoate: residual oviposition repellent activity. J Med Entomol 47:329–337PubMedView ArticleGoogle Scholar
- Seenivasagan T, Sharma KR, Sekhar K, Ganesan K, Prakash S, Vijayaraghavan R (2009) Electroantennogram, flight orientation, and oviposition responses of Aedes aegypti to the oviposition pheromone n-heneicosane. Parasitol Res 104:827–833PubMedView ArticleGoogle Scholar
- Packer MJ, Corbet PS (1989) Size variation and reproductive success of female Aedes punctor (Diptera: Culicidae). Ecol Entomol 14:297–309View ArticleGoogle Scholar
- Lyimo EO, Koella JC (1992) Relationship between body size of adult Anopheles gambiae s.l. and infection with the malaria parasite Plasmodium falciparum. Parasitology 104:233–237PubMedView ArticleGoogle Scholar
- Chambers GM, Klowden MJ (2001) Age of Anopheles gambiae Giles male mosquitoes at time of mating influences female oviposition. J Vector Ecol 26:196–201PubMedGoogle Scholar
- Klowden MJ, Russell RC (2004) Mating affects egg maturation in Anopheles gambiae Giles (Diptera: Culicidae). J Vector Ecol 29:135–139PubMedGoogle Scholar
- Shelton RM (1972) The effects of blood source and quantity on production of eggs by Culex salinarius Coquillett (Diptera: Culicidae). Mosq News 31:31–37Google Scholar
- Downe AERNLG, West AS (1963) The influence of size and source of blood meals on rate of digestion of vertebrate serum proteins in mosquitoes (Diptera: Culicidae). J Kans Entomol Soc 36:200–206Google Scholar
- Downe AE, Archer JA (1975) The effects of different blood-meal sources on digestion and egg production in Culex tarsalis Coq. (Diptera: Culicidae). J Med Entomol 12:431–437PubMedView ArticleGoogle Scholar
- Mather TN, DeFoliart GR (1983) Effect of host blood source on the gonotrophic cycle of Aedes triseriatus. Am J Trop Med Hyg 32:189–193PubMedGoogle Scholar
- Olayemi IK, Ande AT, Danlami G, Abdullahi U (2011) Influence of blood meal type on reproductive performance of the Malaria vector Anopheles gambiae s.s. (Diptera: Culicidae). J Entomol 8:459–467View ArticleGoogle Scholar
- Sumba LA, Okoth K, Deng AL, Githure J, Knols BG, Beier JC et al (2004) Daily oviposition patterns of the African malaria mosquito Anopheles gambiae Giles (Diptera: Culicidae) on different types of aqueous substrates. J Circadian Rhythms 2:6PubMed CentralPubMedView ArticleGoogle Scholar
- McCrae AWR (1983) Oviposition by African malaria vector mosquitoes. I. Temporal activity patterns of caged, wild-caught, freshwater Anopheles gambiae Giles sensu lato. Ann Trop Med Parasitol 77:615–625PubMedGoogle Scholar
- Fritz ML, Huang J, Walker ED, Bayoh MN, Vulule J, Miller JR (2008) Ovipositional periodicity of caged Anopheles gambiae individuals. J Circadian Rhythms 6:2PubMed CentralPubMedView ArticleGoogle Scholar
- Garrett-Jones C, Boreham PFL, Pant CP (1980) Feeding habits of anophelines (Diptera:Culicidae) in 1971–78, with reference to the human blood index: a review. Bull Entomol Res 70:165–185View ArticleGoogle Scholar
- (2013) Inference for a single Proportion: Comparing to a Known Proportion. http://www.stat.ubc.ca/~rollin/stats/ssize/b1.html. Accessed 16 Jan 2015
- (2015) How2stats. http://www.how2stats.net/2011/06/testing-difference-between-correlated.html. Accessed 16 Jan 2015
- IBM-Corp (2011) IBM SPSS Statistics, 20.0 edn. IBM SPSS Statistics for Windows, Armonk, NYGoogle Scholar
- R Development Core Team (2011) R: A language and environment for statistical computing. R Foundation for Statistical computing, ViennaGoogle Scholar
- Rejmankova E, Higashi R, Grieco J, Achee N, Roberts D (2005) Volatile substances from larval habitats mediate species-specific oviposition in Anopheles mosquitoes. J Med Entomol 42:95–103PubMedView ArticleGoogle Scholar
- Huang J, Walker ED, Giroux PY, Vulule J, Miller JR (2005) Ovipositional site selection by Anopheles gambiae: influences of substrate moisture and texture. Med Vet Entomol 19:442–450PubMedView ArticleGoogle Scholar
- Huang J, Walker ED, Otienoburu PE, Amimo F, Vulule J, Miller JR (2006) Laboratory tests of oviposition by the African malaria mosquito, Anopheles gambiae, on dark soil as influenced by presence or absence of vegetation. Malar J 5:88PubMed CentralPubMedView ArticleGoogle Scholar
- Otienoburu PE, Bayoh N, Gimnig J, Huang J, Walker ED, Otieno MF et al (2007) Anopheles gambiae (Diptera: Culicidae) oviposition as influenced by type of water infused into black and red soils of western Kenya. Trop Insect Sci 27:2–5View ArticleGoogle Scholar
- Charlwood JD, Jones MDR (1979) Mating behaviour in the mosquito, Anopheles gambiae s.l. I. Close range and contact behaviour. Physiol Entomol 4:111–120View ArticleGoogle Scholar
- Verhoek BA, Takken W (1994) Age effects on the insemination rate of Anopheles gambiae s.l. in the laboratory. Entomol Exp Appl 72:167–172View ArticleGoogle Scholar
- Kweka EJ, Owino EA, Mwang’onde BJ, Mahande AM, Nyindo M, Mosha F (2011) The role of cow urine in the oviposition site preference of culicine and Anopheles mosquitoes. Parasit Vectors 4:184PubMed CentralPubMedView ArticleGoogle Scholar
- Munga S, Minakawa N, Zhou G, Barrack OO, Githeko AK, Yan G (2005) Oviposition site preference and egg hatchability of Anopheles gambiae: effects of land cover types. J Med Entomol 42:993–997PubMedView ArticleGoogle Scholar
- Haddow AJ, Ssenkubuge Y (1962) Laboratory observations on the oviposition-cycle in the mosquito Anopheles (Cellia) gambiae Giles. Ann Trop Med Parasitol 56:352–355PubMedGoogle Scholar
- Qiu YT, van Loon JJ, Takken W, Meijerink J, Smid HM (2006) Olfactory coding in antennal neurons of the malaria mosquito, Anopheles gambiae. Chem Senses 31:845–863PubMedView ArticleGoogle Scholar
- Marchand RP (1985) A new cage for observing mating behaviour of wild Anopheles gambiae in the laboratory. J Am Mosq Control Assoc 1:234–236PubMedGoogle Scholar
- Lyimo EO, Takken W (1993) Effects of adult body size on fecundity and the pregravid rate of Anopheles gambiae females in Tanzania. Med Vet Entomol 7:328–332PubMedView ArticleGoogle Scholar
- Hogg JC, Thomson MC, Hurd H (1996) Comparative fecundity and associated factors for two sibling species of the Anopheles gambiae complex occurring sympatrically in The Gambia. Med Vet Entomol 10:385–391PubMedView ArticleGoogle Scholar
- Suleman M (1990) Intraspecific variation in the reproductive capacity of Anopheles stephensi (Diptera: Culicidae). J Med Entomol 27:819–828PubMedView ArticleGoogle Scholar
- Beier MS, Beier JC, Merdan AA, El Sawaf BM, Kadder MA (1987) Laboratory rearing techniques and adult life table parameters for Anopheles sergentti from Egypt. J Am Mosq Control Assoc 3:266–270PubMedGoogle Scholar
- Kenawy MA (1991) Development and survival of Anopheles pharoensis and An. multicolor from Faiyum, Egypt. J Am Mosq Control Assoc 7:551–555PubMedGoogle Scholar
- Christophers SR (1960) Aedes aegypti (L.) the yellow fever mosquito, its life history, bionomics and structure. Cambridge University Press, London, NW, New York, NYGoogle Scholar
- Chadee DD, Corbet PS (1991) The gonotrophic status of female Aedes aegypti (L.) overnight at the oviposition site (Diptera: Culicidae). Ann Trop Med Parasitol 85:461–466PubMedGoogle Scholar
- Chadee DD, Corbet PS (1993) The gonotrophic status and diel pattern of entry to outdoor oviposition sites of female Aedes aegypti (L.) (Diptera: Culicidae). Ann Trop Med Parasitol 87:263–268PubMedGoogle Scholar
- Ogbunugafor CB, Sumba L (2008) Behavioral evidence for the existence of a region-specific oviposition cue in Anopheles gambiae s.s. J Vector Ecol 33:321–324PubMedView ArticleGoogle Scholar
- Chen H, Fillinger U, Yan G (2006) Oviposition behaviour of female Anopheles gambiae in western Kenya inferred from microsatellite markers. Am J Trop Med Hyg 75:246–250PubMedGoogle Scholar
- Hurlbert SH (1984) Pseudoreplication and the design of ecological field experiments. Ecol Monogr 54:187–211View ArticleGoogle Scholar
- Kramer WL, Mulla MS (1979) Oviposition attractants and repellents of mosquitoes: oviposition responses of Culex mosquitoes to organic infusions. Environ Entomol 8:1111–1117View ArticleGoogle Scholar
- Huang J, Walker ED, Vulule J, Miller JR (2007) The influence of darkness and visual contrast on oviposition by Anopheles gambiae in moist and dry substrates. Physiol Entomol 32:34–40View ArticleGoogle Scholar
- Overgaard HJ (2007) Effect of plant structure on oviposition behavior of Anopheles minimus s.l. J Vector Ecol 32:193–197PubMedView ArticleGoogle Scholar
- Sumba LA, Ogbunugafor CB, Deng AL, Hassanali A (2008) Regulation of oviposition in Anopheles gambiae s.s.: role of inter- and intra-specific signals. J Chem Ecol 34:1430–1436PubMedView ArticleGoogle Scholar
- Balestrino F, Soliban SM, Gilles J, Oliva C, Benedict MQ (2010) Ovipositional behavior in the context of mass rearing of Anopheles arabiensis. J Am Mosq Control Assoc 26:365–372PubMedView ArticleGoogle Scholar
- Munga S, Minakawa N, Zhou G, Barrack OO, Githeko AK, Yan G (2006) Effects of larval competitors and predators on oviposition site selection of Anopheles gambiae sensu stricto. J Med Entomol 43:221–224PubMedView ArticleGoogle Scholar
- Sileshi G (2006) Selecting the right statistical model for analysis of insect count data by using information theoretic measures. Bull Entomol Res 96:479–488PubMedView ArticleGoogle Scholar
- O’Hara RB, Kotze DJ (2010) Do not log-transform count data. Meth Ecol Evol 1:118–122View ArticleGoogle Scholar
- Fagerland MW, Sandvik L (2009) The Wilcoxon–Mann–Whitney test under scrutiny. Stat Med 28:1487–1497PubMedView ArticleGoogle Scholar
- Zimmerman DW (1998) Invalidation of parametric and nonparametric statistical tests by concurrent violation of two assumptions. J Exp Educ 67:55–68View ArticleGoogle Scholar
- The effect of selection of samples for homogeneity on Type 1 error rate. http://interstat.statjournals.net/YEAR/2001/abstracts/0104002.php
- Seavy NE, Quader S, Alexander JD, Ralph CJ (2005) Generalized linear models and point count data: statistical considerations for the design and analysis of monitoring studies. In: Ralph CJ, John C, Rich TD (eds) Proceedings of the Third International Partners in Flight Conference; 2005; Asilomar, California. Bird Conservation Implementation and Integration in the Americas, pp 744–753Google Scholar