Water source most suitable for rearing a sensitive malaria vector, Anopheles funestus in the laboratory

Background: The insecticide susceptibility status of Anopheles funestus, one of the main malaria vectors in the Afrotropical regions, remains under-studied due to the difficulty of working with this mosquito species. Collecting their larvae in natural breeding sites, rearing and maintaining them in normal laboratory conditions have been a difficult task. Forced-egg laying technique has been a very good tool to generate eggs from adult mosquitoes collected from the wild but rearing these eggs to obtain satisfying portion as adults has always been the problem. In this study, we optimized the development of mosquito species larvae under standard laboratory conditions for desired production of adult mosquitoes that can be useful for insecticide susceptibility tests. Methods: A forced-egg laying technique was used to obtain eggs from gravid female Anopheles funestus collected from Kpome locality in Benin. Eggs were reared in three different water samples (water from the borehole, and two mineral water namely FIFA and Possotômè) and larvae were fed with TetraMin baby fish food. The physico-chemical parameters of the waters were investigated prior to use for egg incubation (introduction of eggs’ batches into water). Results: In contrast to mineral water that had no contamination, the borehole water source was contaminated with lead (2.5mg/L) and nitrate (118.8mg/L). Egg hatching rates ranged as 91.9 ± 4.4%, 89.1 ± 2.5% and 87.9 ± 2.6% in FIFA, Possotômè and borehole water respectively. High emergence of larvae to adult mosquitoes was recorded as in FIFA (74.3%) and Possotômè (79.5%) water. No adult mosquito was obtained from larvae reared in borehole water. Conclusions: This study gave insight on the water sources that could be good for rearing to mass produce An. funestus in the laboratory. More analysis with other local mineral water sources in our environments could be considered in the future, hopefully giving better outputs.


Introduction
Anopheles funestus remains a main malaria vector, and is thereby also responsible for malaria morbidity and mortality in Sub-Saharan Africa 1 . Breeding of this mosquito species like other mosquitoes also requires an aquatic environment where larvae emerge to adult mosquitoes. Water is an important component of the ecosystem of this insect, and the quality of the water is an important determinant in egg laying, for adequate growth Amendments from Version 1 Introduction 1. We have made some minors changes: "at" was replaced by "from". 2. Some sentences were reformulated.
• From "Insecticide susceptibility …. the laboratory". • From "Unfortunately …. malaria vector". • From "Sometimes …... high sensitivity". • The end of the introduction has been re-written for a better understanding. • The word "insecticide" has been introduced in appropriate place in the sentence.

Material and methods
• The small cups used in this study were described; • Details have been given in the revised manuscript to explain what the word Calavi is. • We have specified that the total number of eggs reared was different between the 3 types of water and that larvae rearing duration is the time from L1 to pupae. • The paragraphs "Data analysis" and "Ethical statement" were also corrected.

Results
1. Some sentences have been corrected: • "in" was replaced by "at" in the sentence "nitrate… borehole water". • "recorded" by "obtained" in the sentence "No adult mosquitoes…borehole water". • "adult stage" by "pupae stage" in the sentence "Rearing of An. funestus…15 days".
1. More details have been given to explain what is meant by "egg hatching rate of all incubation days" and "eggs incubation". 2. Statistics were added to show if larval mortality and emerged adults were statistically different between the two mineral waters.

Discussion
1. The word "was" was removed from the sentence "These pH…laboratory conditions". 2. The following sentences have been reformulated: • "This research.…egg shells".
1 Also, "larvae mortality" has been replaced by "larval mortality" throughout the revised manuscript.

Figure 1
The quality of pictures of Figure 1 has been improved.

REVISED
and development from larval stages until adults [2][3][4][5] . An. funestus, unlike the other known malaria vector, An. gambiae, in Sub-Saharan Africa breeds in natural/artificial permanent and semi-permanent water bodies with floating or emerging vegetation like edges of swamps, in weedy and grassy parts of rivers, streams, furrows, ditches and ponds with low salinity and little richness in organic matter 6 . There are reports that Anopheles mosquitoes breed in clear waters with temperatures between 22°C and 32°C 7 . However, decreased oxygen levels caused by water flow and flooding are always responsible for physical damage of mosquito larvae 8 . In addition, breeding water with a pH range of 6.08 to 7.02 is good for weakening the egg shell, so that first instar larvae can emerge 9 . Generally, these chemical properties of larval habitat, including ammonia, nitrate and sulphate concentration, influence larval development and their aquatic survival 3 . Most experiments that study the biology of Anopheles mosquitoes, such as assessments of insecticide susceptibility, use laboratory reared colonies 10 . It is therefore crucial to better understand suitable conditions for rearing of field collected mosquitoes. An. funestus is a difficult mosquito species to handle: not only because its larvae are rarely found during field survey, but also because of their inability to survive in normal laboratory conditions 11 . Since Anopheles funestus represents an important malaria vector across Sub-Saharan Africa 12-15 , it is therefore necessary to find all means to study this malaria vector. Forced-egg laying technique has been a very helpful tool to generate a first filial generation of An. funestus mosquitoes from adults collected on the field 16 . As much as this field tool has been of help, obtaining the desired quantity of An. funestus when rearing its larvae under laboratory conditions for insecticide susceptibility testing has been a big challenge. Sometimes, we recorded high mortality rate even when An. funestus mosquitoes are kept under recommended laboratory conditions. This observation prompted us to rear An. funestus larvae generated from forced-egg laying technique with different water sources but under the same laboratory conditions. Therefore, the aim of this study was to determine the most suitable water source which will obtain the best quantity of F 1 An. funestus for laboratory experiments.

Mosquito collection
Blood fed An. funestus resting indoors were collected in selected rooms at Kpome, a village (6°55′N, 2°19′E) located in the South of Benin. Collection was carried out between 06:00 and 10:00 am using electric aspirators. The collection period corresponds to the dry season in southern Benin when An. funestus densities are likely to be higher. Mosquitoes collected were morphologically identified as An. funestus group using the key of Gillies and Meillon. (1968) 6 , kept in small cups and transferred to the laboratory (Insectariums of the International Institute of Tropical Agriculture of Benin) for rearing of the F 0 to produce the F 1 generation.

Mosquito rearing in the laboratory
In the insectary (T=25°C, RH=70-80% and L/D=12:12), blood fed, semi-gravid and gravid females were kept in the small paper cups (Diameter: 7.3 cm; height: 7.8 cm; capacity: 20cl) for 5 to 6 days after collection until fully gravid stage. Gravid females were then introduced individually and gently into 1.5ml Eppendorf tubes containing cotton soaked in water and surmounted by a filter paper (Wattman 3 mm/1 cm diameter) (Figure 1) 16 . Each Eppendorf tube was checked daily to identify females of An. funestus that have laid eggs, mosquitoes were gently removed from the tubes and transferred to a new Eppendorf tubes containing cotton and silica gel and stored at -20°C for subsequent experiments. Twenty-four hours post-oviposition, eggs from a single mosquito were divided into 3 batches and were allowed to hatch in small cups for larvae emergence, which were later transferred into rearing bowls containing 3 different types of water (Borehole water collected in Calavi, a southern Benin locality, and local mineral waters named FIFA and Possotômè) ( Figure 1). Water of each larvae bowl was changed every two days to reduce mortality and larvae were fed daily with Tetramin™ baby fish food. The larvae were monitored daily during the four larval stages of the mosquito up to the adult stage (F1 An. funestus). Each experiment was repeated at least 8 times with each type of water. The growth and development yield was evaluated based on: (i) eggs hatching duration, (ii) larvae rearing duration corresponding to duration from L1 stage till the first pupae stage, (iii) larval mortality rate (iv) adult emergence rate.
Sheets were established to collect data manually on these parameters, such as the number of eggs incubated in each type of water, the number of hatched eggs, larval and pupae mortality and number of daily emerged adult mosquitoes.
Physico-chemical parameters of breeding water Each water sample was analyzed for physico-chemical parameters in the laboratory of water and food quality of Agriculture Environment and Health (AgroEcoHealth) platform of IITA-Benin. Temperature and pH of each water sample were determined using the pH meter WAG-WE30200 (Wagtech Projects, Berkshire, UK). Conductivity and Total dissolved solid (TDS) were also determined using the conductivity/TDS meter WAG-WE30210 (Wagtech Projects, Berkshire, UK). Before analysis, electrodes of these materials were sensitized, calibrated and rinsed with deionized water. Water samples were then analyzed, and all parameters were read and recorded.
The quantity of nitrate, nitrite and chloride was determined using the Photometer 7100: WAG-WE10441 (Wagtech Projects, Berkshire, UK). Three replicates of each water sample were introduced into beakers and the reagents were added as recommended by the manufacturer (Wagtech Projects, Berkshire, UK). The mixture was incubated to stand for 10 min (Nitrate and nitrite) and for 2 min (chlorine) to allow the appearance of color. Each beaker was then inserted into the photometer and concentrations were directly displayed and recorded.
Calcium and fluorine were quantified using the W-22XD.23XD HORIBA multi-probe (HORIBA, ltd Japan). The electrodes were also calibrated and rinsed with sterile deionized water and then with the water samples to be analyzed. The quantities of calcium and fluorine were recorded after homogenization of the samples.
Heavy metals (Cadmium, Lead and Copper) were quantified using METALYSER HM 3000 (TRACE2O, Berkshire, UK) by the reverse voltammeter method. The electrodes were placed and conditioned according to the desired metal as recommended by the manufacturer. Heavy metal was quantified in 70 ml of water with appropriate reagents (buffer and standard) according to the desired metal. After 5 mins of incubation, the concentration of the metal and the corresponding graphs (in the form of a peak) were displayed on a tablet connected to the machine.

PCR-based species identification
All females used for individual oviposition were identified using PCR as belonging to the An. funestus group. DNA was extracted from a total of 94 mosquitoes using the Livak protocol 17 , followed by PCR species identification using the protocol described by Koekemoer et al. (2002) 18 .

Data analysis
Data were inserted on excel sheets and analyzed using SPSS 17.0. Chi square test was used to analyze the difference between hatching rate in different water samples. The easy to use online software (Fischer exact test) was used to assess the difference in larval mortality rate and adult emergence rate according to the water samples. The significance level was set at 5%.

Ethical statement
The request for ethical approval was not applicable for this study, according to the International Institute of Tropical Agriculture (IITA) Ethical Committee (IITA, 08 P.O. Box 0932, Tri-Postal, Cotonou, Benin). However, there was a focus group discussion with the community and household heads where verbal and written consent was obtained for mosquito collections in the community after the study aims and objectives were explained. Since mosquitoes were collected using electrical aspirators, no insecticide spray or human bait methods were used for mosquito collections during this study.

Species identification
Molecular identification of 94 females used for forced-eggs technique revealed that they all belonged to An. funestus s.s.

Physico-chemical parameters of water samples
The physico-chemical parameters of the different water types used for An. funestus rearing are summarized in Table 1. These results showed that Possotômè mineral water had a pH of 7.9 whereas the pH obtained with FIFA mineral water and borehole water were 5.9 and 6.2, respectively. The total dissolved solids in Possotômè water (386 mg/l) was high compared to borehole water (131mg/l) and FIFA water (29.9mg/l). The same trend was observed with the total conductivity characterized by a high quantity of minerals in Possotômè water (775 μS/cm), which was significantly different to FIFA water (60 μS/cm) and borehole water (265 μS/cm) (p <0.05). Indeed, calcium and chloride concentration were higher in Possotômè mineral water (54 mg/l calcium, 110 mg/l chloride) than in FIFA There was no significant difference in the egg hatching duration between the borehole water (5 days) and mineral waters (4 days) (P = 0.0722). The eggs hatching rate of all repetitions (experiments) of each water sample are summarized in Table 2.
These hatching rates varied between incubation days (the day the eggs were introduced in the water sample) and ranged from 12.61 ± 0.11% to 41.21 ± 6.11% for borehole, 16% to 55.57 ± 6.46% for FIFA and from 19.6 ± 2.38% and 43.77 ± 5.01% for Possotômè. No significant difference was found in the daily hatching rates for borehole (p = 0.0637), FIFA (p = 0.1450) and Possotômè waters (p = 0.080). Overall, no significant difference in eggs hatching rate was observed between the three waters (FIFA 91.9 ± 4.4%, Possotômè 89.1 ± 2.5% and borehole 87.9 ± 2, 6%) (P<0.05). Larval mortality rates obtained were respectively, 97.36%, 17.5% and 14.06% in borehole, Possotômè and FIFA waters (Table 3). There was a significant difference in larval mortality with borehole water and mineral waters (p<0.05). No significant difference was observed in larval mortality between FIFA and Possotômè mineral waters (P = 0.3573). The percentage of adult mosquitoes that emerged from FIFA and Possotômè mineral waters were respectively 74.36% and 79.50%. No adult mosquitoes were obtained from borehole water (Table 3). Another observation was that the rate of emerged adults in Possotômè was slightly higher than in FIFA mineral water but not significant (P = 0.1823). Rearing of An. funestus larvae to adults with both mineral waters took about 10 days, while for borehole water, rearing of larvae to pupae stage took as long as 15 days.

Discussion
Water quality is an important factor for oviposition of the female mosquito, and also influences the emergence of adult mosquitoes from larvae stages 4 . However, physico-chemical parameters such as temperature, pH, dissolved oxygen, nitrate and sulphate concentrations are likely to affect the development and survival of mosquito larvae 3 . The different values of physico-chemical parameters obtained in this study could give a better understanding on the environmental requirements needed to produce good yield of F 1 An.funestus mosquito in the laboratory from the ones collected in the wild. The pH ranges (5.9 to 7.9) recorded in the different water sources could be considered suitable for mosquito breeding. These pH were similar to that found in breeding water samples, with pH values ranging from 4.0 to 7.8 considered favorable for normal development of An. gambiae under laboratory conditions 9,19 . pH values recorded in all sampling waters might not have an effect on the eventual yield of An. funestus in this study. A previous study also revealed that mosquito larvae can thrive well in water with neutral or slightly alkaline pH 1,20 . The temperature of water samples (25°C) used in this study is known to be suitable for An. funestus mosquito breeding 7 .
This research also showed similar high hatching rates of eggs with FIFA, Possotômè and borehole waters, indicating that physicochemical compositions of the different water samples do not influence the weakening of An. funestus egg shells. However, a high larval mortality rate (97%) was observed with borehole water compared to mineral water samples that produced good emergence rates. High larval mortality rate recorded could be attributed to the high nitrate dose in borehole water (118.8mg. l). This nitrate concentration in this water sample is higher than the maximum limit of 50mg/l nitrate dose authorized for human consumption 21 , and could also be toxic for mosquito larvae. The main toxic action of nitrate on aquatic animals is the conversion of oxygen-bearing pigments (hemoglobin and hemocyanin) into the inhibited forms (methaemoglobin) which are not able to fix and carry oxygen 22-24 . Therefore, the lack of oxygen can cause the death of the mosquito larvae. The high nitrate content in the borehole water could be justified by the phreatic origin of this water, because ground water has always been described to contain toxic nitrate concentrations exceeding the standard 9,21 . Water sources that should be used for the breeding of aquatic stages of An. funestus should not have similar physico-parameters as ground water. Nitrate and phosphate residues derived from chemical fertilizers used in agriculture are other potential pollutants that could be found in aquatic waters 25   This manuscript try to describe some qualities needed for water used for rearing Anopheles funestus mosquito species in the laboratory. The study deals with an important issue all medical entomologist faces to when trying to establish a lab stain of mosquito. Results obtained this Anopheles funestus manuscript of Tchigossou are very interesting and very informative. However even if the manuscript et al and data presented are interesting, there are some minor concerns I will like the authors to take into consideration so that their work will be more efficient.
: Comments 1) In the section Material and methods, it would certainly be interesting for authors to give an approximate size of the small cup they used to keep their mosquitoes after identification. This aspect could be important as the manuscript try to present some interesting condition for the rearing of Anopheles mosquito in the laboratory funestus 2) In the text (mosquito rearing in the laboratory) authors speak about "Calavi". I think they should give more details on what is Calavi. Is not easy for the reader with no information about this locality to know what Calavi is. 3-In the "mosquito rearing" section, it would have been interesting for authors to be precise that the total number of eggs reared was different or the same between the 3 types of water. 4-As the present manuscript is trying to describe an experimental design to improve the rearing of in the laboratory, I think authors should ameliorate the quality of the pictures in figure Anopheles funestus 1. This will be helpful for someone who will try to reproduce what was done in the present study. Also it could be interesting for the reader if authors could give more details on "eggs incubation" 5-The two last sentences of the results section (Rearing yield of wild population of ) seem to An. funestus be contradictory. Indeed, on one side authors said "No adult mosquitoes were recorded from borehole water", on another side they said "… for borehole water, rearing of larvae to adult stage took as long as 15 days. It's a bit difficult for the reader to understand how authors did to estimate these 15 days to obtain adult from borehole water, when it's said by authors that with this water no adult mosquitoes were recorded? So maybe authors could clarify this last part of the "results" section.

If applicable, is the statistical analysis and its interpretation appropriate? Yes
Are all the source data underlying the results available to ensure full reproducibility? Yes 5) "eggs incubation" means introduction of the eggs batches into water. As per Reviewer's suggestion, this precision has been given in the revised manuscript. 6) We agree with the Reviewer's comment. Larvae reared in this water sample reached to pupae stage but no emergence was recorded. So, for this water sample "adult stage" was replaced by Introduction 1) ... for adequate growth and until adults development from larval stages 2) Insecticide susceptibility studies are important for checkmating the effectiveness of control tools used to control malaria vectors . Achieving success in this field requires a good understanding of the bio-ecology of the vector. This will guide us to organize suitable conditions to maintain mosquito samples collected from fields in the laboratory.
For me this transition is not appropriate. I suggest the authors to say rather that "most of Comment: experiments to study the biology of Anopheles mosquitoes, such assessment of insecticide susceptibility, use laboratory reared colonies. Is this therefore crucial to better understand suitable conditions for rearing of field collected mosquitoes".
3) Unfortunately, there are reports that is involved in malaria transmission across Africa . An. funestus It is therefore necessary to find all means to study this malaria vector.
I suggest the authors to say rather "Since Anopheles funestus represents an important Comment: malaria vector across Sub-Saharan Africa, it is therefore necessary to find all means to study this malaria vector." 4) ........... for susceptibility testing....... insecticide 10 5) Sometimes, we also have the problem of recording a high mortality and low hatching rate of field mosquitoes under suitable laboratory conditions because of their high sensitivity.
This sentence is not clear. Can the author reformulate it? Question:

6)
I found the end of the introduction laborious from " Sometimes, we also have the Comment: problem..... " to the end Material and methods 7) larvae rearing duration corresponding to duration from oviposition till the first pupae stage Most of the time, larvae rearing duration is the time from L1 to Puape Comment:

8)
Larval mortality, not larvae mortality Comment: Data analysis 9) Chi square test was performed to analyze the difference in hatching rate between different water samples and hatchingduration according to the water types.
I guess that hatching duration was expressed in days and that the authors rather compared Comment: means. I don't understand why the chi square test was used.
Ethical statement 10) Mosquitoes were collected in the morning using electrical aspirators. No insecticide spray or human bait methods were used for mosquito collections during this study.
I suggest the authors to say rather "Since mosquitoes were collected using electrical Comment: aspirators, no insecticide spray or human bait methods were used for mosquito collections during this study."

11)
Nitrate was almost absent in both mineral water (Possotome and FIFA) but was present in high concentration of 118.8mg/l in borehole water (Table I).
: Nitrate was almost absent in both mineral water (Possotome and FIFA) but was present Comment at high concentration of 118.8mg/l in borehole water (Table I).

12)
I don't understand what the authors mean by egg hatching rate of all Comment and question: incubation days. Could they please explain or use a more clear expression?
13) There was a significant difference in larvae mortality with borehole water and mineral waters (p<0.05).
The authors didn't mention if larval mortality was statistically different or not between the two Comment: mineral waters 14) No adult mosquitoes were recorded from borehole water (Table 3).
No adult mosquitoes were from borehole water (Table 3).

Comment: obtained
15) Another observation was that the rate of emerged adults with Possotome was slightly higher when compared to FIFA mineral water.
Please add statistics Comment: Please add statistics Comment: Discusssion 16) These pH were similar to that found in breeding water samples, with pH values ranging from 4.0 was to 7.8 considered favorable for normal development of under laboratory conditions An. gambiae Remove "was" Comment: 17) This research also showed high hatching rates of eggs with FIFA, Possotome and borehole similarly waters, indicating that physico-chemical compositions of the different water samples do not influence the weakening of egg shells An. funestus Say rather "similar high..." or "comparable high...." Comment:

18)
Can the authors use the term "larval mortality" rather than "larvae mortality"? Question: 19) This nitrate concentration in this water sample is higher than the maximum limit of 50mg/l nitrate dose authorized for human consumption and could also be toxic for the larvae development.
, Say rather "This nitrate concentration in this water sample is higher than the maximum limit of Comment: 50mg/l nitrate dose authorized for human consumption . , and could also be toxic mosquito larvae

If applicable, is the statistical analysis and its interpretation appropriate? Partly
Are all the source data underlying the results available to ensure full reproducibility? Yes

Are the conclusions drawn adequately supported by the results? Yes
No competing interests were disclosed.

Competing Interests:
I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. 15) We agree with Reviewer's comment. We have now corrected the sentence and added statistics to this statement in the revised manuscript (P = 0.1823).

16)
As per Reviewer's suggestion, we have removed 'was' from the sentence in the revised manuscript. Please see first paragraph of the section Discussion 17) As per Reviewer's suggestion we have now corrected the sentence in the revised manuscript.
, Please see in the section Discussion 18) We agree with the Reviewer's comment. "larvae mortality" has been replaced by "larval mortality" throughout the revised manuscript. 19) We agree with the Reviewer's comment. The comment has been addressed in the manuscript and the sentence has been modified as suggested.
The authors have nothing to declare as far as the conflict of interest is Competing Interests: concerned