Is the expression of autogeny by Culex molestus Forskal (Diptera: Culicidae) influenced by larval nutrition or by adult mating, sugar feeding, or blood feeding?

ABSTRACT:

Culex molestus Forskal is suspected to have been introduced into southern Australia during the 1940s. Investigations to determine factors influencing the expression of autogeny, the response of this mosquito to potential blood meals, and the subsequent influence on oviposition were undertaken. Immature mosquitoes raised at five feeding regimes had mortality rates, development rates, wing length, and autogenous egg raft size measured. All surviving female mosquitoes laid autogenous eggs but there was a significant difference between the mean number of eggs per raft. For mosquitoes raised at each of the feeding regimes, there was a significant linear relationship between the number of eggs per autogenous egg raft and wing length. Newly emerged mosquitoes were offered a blood meal (i.e., rodent) daily but no blood feeding occurred until the autogenous egg raft was laid. There was no statistical difference in the rate of autogenous oviposition or post-oviposition blood feeding between control or treatment groups. The results of this study indicate that Cx. molestus is perfectly adapted to subterranean habitats in close association with human habitation, but their preference to delay blood feeding until up to day 8 following emergence may reduce their relative importance as a vector of arboviruses.

INTRODUCTION

Culex molestus Forskal belongs to the Culex pipiens subgroup of mosquitoes that includes a number of globally important vectors of disease-causing pathogens. It was first described from specimens collected in Egypt during the late 1700s (Knight and Abdel Malek 1951) and has been recorded from many countries, although there is ongoing debate surrounding the taxonomic relationship of this species within the Culex pipiens, in particular the genetic differences between Culex pipiens form pipiens and Culex pipiens form molestus which exhibit both biological and ecological differences (Vinogradova 2000). The species was thought to have been introduced into to Australia during World War II with movement of U.S. defense forces through Melbourne in the early 1940s (Dobrotworsky and Drummond 1953, Drummond 1951, Lee et al. 1989). The species is currently known to occur in New South Wales, Victoria, Tasmania, South Australia, and Western Australia, where it is typically associated with subterranean habitats (e.g., septic tanks) in urban environments.

While mosquito-borne disease is a public health concern in Australia and a large number of local and introduced mosquito species have been associated with the transmission of disease-causing pathogens (Russell and Kay 2004), little attention has been paid to Cx. molestus. However, the species has been found to be an effective laboratory vector of arboviruses, including Murray Valley encephalitis virus (McLean 1953, Anderson 1954), and Ross River virus (RRV) and Barmah Forest virus (BFV) have been isolated from field collected specimens (Vale et al. 1985). An understanding of the biology and ecology of Cx. molestus will greatly assist in assessing the risk this species poses to the public from the transmission of endemic and/or exotic arboviruses.

Culex molestus is an autogenous species (Dobrotworsky 1954, Marshall 1944, Dobrotworsky and Drummond 1953). Autogeny is the ability of female mosquitoes to develop their first batch of eggs without a reliance on a blood meal (Clements 1992), and laboratory studies strongly suggest that Cx. molestus displays obligatory autogeny due to the relatively short period between the emergence of females and the oviposition of autogenous egg rafts as well as the high rate of autogenous egg rafts laid (Kassim et al. 2011). The documented ecological niche of Cx. molestus, where immature stages are most commonly collected from subterranean or enclosed water holding containers such as septic tanks (Dobrotworsky 1965, Mohsen et al. 1995), suggests that the species is less likely to encounter potential vertebrate blood meals. However, it has not been documented if the species will take a blood meal, if available, prior to laying the first batch of eggs and how that blood meal may influence the expression of autogeny and/or additional anautogenous egg laying.

The role of mating in autogenous and anautogenous egg development is crucial, with a single mating providing a female mosquito with sufficient sperm to fertilize the eggs of numerous clutches (Craig 1967). While O'Meara (1985) stated that the mating process might occur either before or after blood feeding for anautogenous egg development, for autogenous mosquitoes, particularly for species that display facultative autogeny, the process of mating may regulate the expression of autogeny. There is evidence that substances from the male accessory gland introduced through the insemination process have the potential to influence a wide range of physiological mechanisms in female mosquitoes including host-seeking and oviposition and egg development (Klowden and Chambers 1991).

In mosquito species that display facultative autogeny, substances from the male accessory gland have been shown to influence the expression of autogeny, with some mosquito species shown to be potentially anautogenous for several days following emergence (O'Meara and Evans 1977). O'Meara and Petersen (1985) stated that the mating process can stimulate the initiation of autogenous egg development in crabhole mosquitoes, such as Deinocerites pseudes (Dyar and Knab) and Deinocerites cancer (Theobald) and that insemination usually occurs between five- to seven-day-old males and females >1 day old. Laboratory studies have shown that sugar-fed De. cancer females had larger autogenous egg rafts when they were either mated or unmated (O'Meara 1979).

Larval nutrition at the immature stages of mosquito life is one of the important factors that influence the expression of autogeny in mosquitoes (Clements 1992, O'Meara 1979). Even though Cx. molestus does not require a blood meal, they still need sufficient larval nutrients for production of the first batch of eggs. Some obligatory autogenous species are able to lay multiple egg batches without blood-feeding if larval reserves are sufficient (O'Meara 1985). The quantity and quality of larval diet for obligatory autogenous species have been shown to be influencing factors in determining fecundity during the first ovarian cycle (O'Meara and Krasnick 1970, Kalpage and Brust 1974).

Sugar-feeding is also one of the factors that can influence the expression of autogeny in mosquito species (O'Meara and Petersen 1985). Su and Mulla (1997) believed that sugar-feeding enhanced the fertility rate in autogenous females in Culex tarsalis (Coquillett). The sugar-feeding also increased by between 2% and 14% the number of autogenous egg rafts by Aedes taeniorhynchus (Wiedeman) raised under poor larval nutrition conditions (Lea 1964).

There is some uncertainty regarding the physiological processes that are completed by autogenous mosquitoes when they take a blood meal during the period of autogenous egg development, with a paucity of published studies on species that display obligatory autogeny. Laboratory studies have shown that, when offered a potential blood meal, most gravid females of Aedes togoi (Theobald) with autogenously developed oocytes declined to feed but those that did feed developed an additional series of follicles (McGinnis and Brust 1985). Sweeney and Russell (1973) also found that females of Anopheles amictus hilli (Woodhill and Lee), a known autogenous species, willingly took a blood meal if available, even though they did not require a blood meal to mature the first egg batch. The additional nutrition may not contribute to increased fecundity if autogenous ovarian development has progressed before the blood feeding occurs (O'Meara and Evans 1973).

The aims of this laboratory investigation were to determine the influence of four factors on the expression of autogeny in Cx. molestus, the role of larval nutrition on the size of autogenous egg rafts, the role of mating in the expression of autogeny, the influence of sugar feeding on the expression of autogeny, and the response of females to the availability of a blood meal prior to oviposition.

MATERIALS AND METHODS

Mosquito rearing

All mosquitoes used in these experiments were from a laboratory colony of Cx. molestus established in 2010 using field-collected specimens from Sydney Olympic Park, NSW (Kassim et al. 2011). Mosquitoes were housed within the insectary of the Department of Medical Entomology, University of Sydney, and Westmead Hospital under animal ethics approval number 8001/04–10 (Colonization and maintenance of mosquito stock colonies). The insectary operates at temperature of 23–26° C, relatively humidity 60–90%, and with a 12:12 h light: dark regime.

Mating and autogeny

Immature stages of mosquitoes were raised under controlled conditions as described by Kassim et al. (2011). Egg rafts were allowed to hatch in small cups (approximately 200 ml) of deionized water before larvae were transferred to trays approximately 35×45×10 cm in size and containing approximately 4.0 liters of deionized water. Larvae were fed a stock solution comprised of 3.5 g brewer's yeast (Brewer Yeast, Healthy Life, Burwood) and 3.5 g fish flakes (Warley's Tropical Fish Food Flakes, Hartz Mountain Corporation, NJ) in 500 ml of deionized water. For 1^st and 2^nd instars, approximately 10 ml of the stock solution was added daily, and, for 3^rd and 4^th instars, 5 ml of the stock solution plus approximately 0.16 g pellet of fish flakes were added daily. Pupae were transferred to smaller cups (approximately 200 ml) and placed in a new cage for emergence.

Two cages (approximately 16×16×16 cm) were used to house adult mosquitoes. In one cage, 100 male and 100 female adult mosquitoes were maintained together, and in the second cage only 100 virgin females were housed. Mosquitoes in both cages were provided access to a cotton pad soaked in a 10% sugar solution. Commencing 72 h after adult mosquito emergence, a sub-sample of 60 female mosquitoes from each cage was collected at 24 h intervals via suction tube and the number of autogenous egg rafts laid was recorded. Eggs were inspected to determine the presence of viable embryos or put aside to record hatching rate. Adult females were dissected and spermathecae inspected to determine the presence of sperm to confirm mating. A comparison was made between the numbers of eggs per raft between mosquitoes in the two cages to determine the role of mating in the expression of autogeny.

Larval nutrition and autogeny

Larvae were provided a diet based on a stock solution comprised of brewer's yeast (Brewer Yeast, Healthy Life, Burwood) and fish flakes (Wardley's Tropical Fish Food Flakes, The Hartz Mountain Corporation, NJ). Larvae were reared under five different food regimes (i.e., D1, D2, D3, D4, and D5) (Table 1).

Table 1.  Rate of food provision to immature stages of Culex molestus under five diet regimes.

Feeding regime	1^st– 2^nd instar stage	3^rd– 4^th instar stage
D1	0.2 mg per larva	0.4 mg per larva
D2	0.4 mg per larva	0.8 mg per larva
D3	0.8 mg per larva	1.6 mg per larva
D4	1.2 mg per larva	2.4 mg per larva
D5	1.6 mg per larva	3.2 mg per larva

For each of the five group food regimes, three replicates of 50 larvae within 2 h of eclosion were placed in white plastic containers (7×10×15 cm) containing approximately 500 ml of deionized water. To determine the influence of larval diet in autogeny expression, the number of live larvae and their instar were recorded daily as well as date of pupation and date of emergence. A one-way ANOVA was used to compare the difference among the mean development times for each immature stage in five group food regimes.

Pupae were removed from their larval rearing containers and, for each food regime, were transferred into emergence cups and placed in a cage (12×12×12 cm) where adults emerged. Adult mosquitoes raised under each of the food regimes were maintained on a 10% sugar solution for five days and then removed and killed by being placed at –20° C for 1 h. All females were dissected under a microscope and the total number of autogenous eggs per female and wing lengths of each female were recorded. Wing length is an indicator of adult mosquito body size (Nasci 1986, Nasci 1990) and was measured from the alular notch to the apical margin, excluding fringe scales. Box plots were used to illustrate the distribution of number of autogenous eggs per raft and wing length resulting from five different diets. Kruskal-Wallis test for non-parametric analysis of variance was used to test for homogeneity of the distribution of this variable by different diets. The Mann-Whitney test was used for multiple pair-wise comparison between diets, and P-values were Bonferonni corrected. To determine the relationship between wing length and autogenous fecundity of Cx. molestus, least squares regression analysis was used on wing length vs fecundity of mosquitoes reared from five food regimes.

Sugar feeding and autogeny

Approximately 100 male and female adult mosquitoes were placed in two different cages (16×16×16 cm). The mosquitoes in one cage were provided access to a cotton pad soaked in 10% sugar solution while the second cage only had access to a cotton pad soaked in water alone. The oviposition container that contained larval water as oviposition water was placed on day five of emergence in each cage. Each day, the total number of autogenous egg rafts laid and eggs per raft for each cage were counted under the microscope and adult females were dissected and spermathecae inspected to determine the presence of sperm. The number of autogenous eggs per raft was compared by T-test analysis and the hatching rate using Chi-square test to determine the significance of any difference in sugar- feeding influence for autogeny expression.

Blood feeding and autogeny

To investigate the response of autogenous mosquitoes to the availability of a potential blood meal, pupae were immediately removed from the immature trays into cages (30×30×30 cm) to ensure that all adults emerged within 24 h. For each of the two cages, treatment and control, approximately 100 adults (approximate ratio 50M: 50F) within 2 h of emergence were placed into cages (16×16×16 cm). The mosquitoes were maintained on a 10% sugar solution and allowed to mate. In the treatment cage, mosquitoes were provided with an anesthetised rat as a blood meal for 45 min daily for five days, and on each day the engorged females (if present) were taken out by suction tube and then transferred into smaller cages (12×12×12 cm). The oviposition container that contained larval water as oviposition water was placed on day 5 in each cage. No blood meals were offered to mosquitoes in the control cage. The total number of engorged females and the total number of egg rafts laid were recorded.

To examine the post-oviposition blood-feeding response of Cx. molestus, the treatment cage continued to be provided with a rat for a blood meal for 45 min each day from day 8 to day 12. Each day, the engorged females were taken out by suction tube and transferred into small cages (12×12×12 cm) and the total number of engorged females and the number of egg rafts laid were recorded. The mean numbers of autogenous egg rafts per day and engorged females per day (post-autogeny) were compared using one-way ANOVA to determine the significance of any difference in blood-feeding response. The oviposition rate of treatment and control cage were compared by Chi-square test to determine the significance of any difference in blood-feeding influence for autogeny expression.

RESULTS

Mating and autogeny

All 60 mated females laid autogenous egg rafts with a mean (± SE) number of eggs per raft of 55.8 ± 6.3 eggs per raft with all eggs hatching successfully. All mated mosquitoes were confirmed by dissection of spermatheca and detection of sperm in 100% of specimens. For unmated females, a total of 44 out of 60 females (73.3%) of mosquitoes laid egg rafts with a mean (± SE) number of eggs per raft laid autogenously of 32.1 ± 10.6 eggs per raft; however, the eggs were not viable.

Larval nutrition and autogeny

There was a highly significant difference (F=158.52, d.f.=14, P<0.001) in the mortality rate of immature mosquitoes raised at the five different food regimes (Figure 1). In particular, D1 had a significantly higher mortality rate than D2 (p<0.001), which in turn had a significantly higher rate (p<0.001) than D3, D4, and D5 which were comparable after Bonferonni adjustment of all pair-wise multiple comparisons. There was no mortality of immature stages raised at D4 and D5. However, the mortality rate increased in D3 to 3.3%, in D2 to 22.7%, and D1 to 38.7%. Within these three groups, mortality was greatest in the 4^th instar and pupae after day 13 of emergence.

Figure 1. Total mortality of Culex molestus immatures raised at five different feeding regimes (n = 50) (P < 0.001, Kruskal-Wallis).

There was a substantial difference in the mean (± SE) development time for each immature stage on five different food regimes, with the longest mean development time of 26.0 ± 1.0 days at D1 and the shortest was 10.7 ± 0.6 days at D5. The mean development times of Cx. molestus larvae raised at five different food regimes recorded during the larval development experiments were 26.0 ± 1.0 days in D1, 23.3 ± 0.6 days in D2, 15.7 ± 0.6 days in D3, 13.7 ± 0.6 days in D4, and 10.7 ± 0.6 days in D5 (Table 2). There was a significant difference in the mean time to completion of the 1^st instar (F= 15.25, d.f. = 4, P < 0.05), 2^nd instar (F= 130.02, d.f. = 4, P < 0.05), 3^rd instar (F= 233.70, d.f. = 4, P < 0.05), 4^th instar (F= 359.70, d.f. = 4, P < 0.05), and pupae (F= 273.79, d.f. = 4, P < 0.05) raised under each of the five food regimes. However, the mean time to completion of each immature stage on each of the food regimes was not consistent (Table 2).

Table 2.  Mean development time of Culex molestus for each immature stage (time from hatch to >50% reach instar stage) in five different food regimes.

	I	II	III	IV	P
^1Values followed by the same letter are not significantly different (p > 0.05), one-way ANOVA.
D1	5.33 ± 0.58a^1	13.33 ±0.58a	19.33 ±0.58a	23.33 ±0.58a	26.00 ± 1.00a
D2	4.33 ± 0.58b	13.33 ± 0.58a	17.33 ± 0.58b	21.33 ± 0.58b	23.33 ± 0.58b
D3	3.67 ± 0.58bc	7.67 ± 0.58b	10.67 ± 0.58c	13.67 ± 0.58c	15.67 ± 0.58c
D4	3.00 ± 0.00cd	6.67 ± 0.58bc	9.67 ± 0.58c	11.67 ± 0.58d	13.67 ± 0.58d
D5	2.33 ± 0.58d	5.33 ± 0.58c	7.67 ± 0.58d	8.67 ± 0.58e	10.67 ± 0.58e

All surviving female mosquitoes (D1: n = 12, D2: n = 23, D3: n = 30, D4: n = 30, and D5: n = 30) raised on each of the feeding regimes laid autogenous eggs and there was a significant difference (χ²= 113.93, d.f.= 4, P < 0.001) between the numbers of eggs per raft of Cx. molestus under five different food regimes (Figure 2). However, the Bonferonni correction test showed that P = 0.620 for D1 and D2 and P < 0.001 for all others pair-wise.

Figure 2. Total number of eggs per autogenous egg raft laid by Culex molestus raised under five different food regimes (P < 0.001, Kruskal-Wallis).

There was a significant difference (χ²= 114.18, d.f.= 4, P < 0.001) between the sizes of wing length per mosquito raised on each of the five different food regimes, and the Bonferonni correction test also revealed that P = 0.020 for D1 and D2 and P < 0.001 for all others pair-wise (Figure 3).

Figure 3. The wing length per mosquito of Culex molestus under five different food regimes (P < 0.001, Kruskal-Wallis).

When comparing the relationship between mean wing length and immature food regime, there was a statistically significant (R = 0.996) interaction between diet and the linear relationship with wing length (F= 43.03, d.f. = 4, P < 0.001). Therefore, the relationship between the number of eggs per autogenous egg raft and wing length for mosquitoes raised at each of the five food regimes was analyzed separately. For each of the five food regimes, there was a significant (P < 0.05) linear relationship between the number of eggs per autogenous egg raft and wing length (Figure 4). However, there were differences in the R² value for each of the food regimes, as well as the coefficient, in the relationship between wing length and autogenous egg batch size for mosquitoes raised on the five food regimes (Table 3).

Figure 4. Relationship between wing length and autogenous fecundity of Cx. molestus under five different food regimes.

Table 3.  Results of regression analysis of mean wing length to number of autogenous egg raft size of Culex molestus raised at five food regimes.

Food regime	p-value (ρ)	R^2 value (r)	Constant (Ω)	Wing length coefficient (b)
D1	0.21	0.92	89.45	37.76
D2	0.16	0.98	126.24	49.96
D3	0.26	0.98	101.37	41.76
D4	0.00	0.93	220.03	77.65
D5	0.05	0.88	301.74	99.90

Sugar feeding and autogeny

There was a significant difference (t = 5.20, d.f. = 58, P < 0.001) between the total number of eggs per raft adult Cx. molestus maintained with sugar and without access to sugar. There was also a significant difference (Fisher's exact test P < 0.001) between the rate of successful hatching between autogenous eggs laid by sugar-fed (98%) and non-sugar-fed (60%) mosquitoes.

Blood feeding and autogeny

No mosquito took a blood meal, or was attracted to an available blood meal, between day 1 and day 5 following emergence. However, mosquitoes began taking blood meals on day 8 post-emergence; approximately 48 h following oviposition of the first autogenous egg rafts (Figure 5). For 50 mosquitoes offered a pre-oviposition blood meal, a total of 46 (92.0% of female mosquitoes in cage) mosquitoes laid autogenous egg rafts and 37 (80% of egg-laying mosquitoes in cage) mosquitoes took a blood meal (Figure 5). For 50 mosquitoes not offered a pre-oviposition blood meal, a total of 48 (96.0% of female mosquitoes in cage) mosquitoes laid autogenous egg rafts and 40 (80% of egg-laying mosquitoes in cage) mosquitoes took a blood meal (Figure 5). There was no significant difference in the rate of oviposition (X²= 0.13, P = 0.72) between the two groups of mosquitoes.

Figure 5. Number of blood-feeding adult Cx. molestus and number of autogenous egg rafts deposited per day post-emergence when offered a pre-oviposition blood meal (top) or no pre-oviposition blood meal (bottom).

DISCUSSION

This study has demonstrated the influence of four contributing factors in the expression of autogeny by Cx. molestus. Our study has shown that the insemination process influences the expression of autogeny in Cx. molestus with both mated and unmated mosquitoes developing autogenous eggs (although eggs laid from unmated females were not viable due to the absence of sperm). However, fewer individuals developed autogenous eggs in the absence of mating. Similarly, O'Meara and Petersen (1985) reported that 13.3% of De. pseudes (Texas strain) and up to 100% of De. cancer (Belize strain) were recorded as having ovarian development in the absence of mating in laboratory trials. A recent laboratory study of Anopheles gambiae sensu stricto (Giles) showed that females inseminated by sterile males (i.e., no active sperm) exhibited normal post-copulatory responses such as blood feeding and laying eggs and, although no eggs resulting from mating with sterile males hatched, there was no significant difference in the mean number of eggs laid by females mated with sterile or fertile male mosquitoes (Thailayil et al. 2011). Similarly, Su and Mulla (1997) found that in laboratory tests, approximately 40% of Cx. tarsalis laid autogenous egg rafts without mating.

Larval nutrition plays an important role in determining the size of autogenous egg rafts (Clements 1992), and a nutrient poor larval diet has been shown to reduce both the penetrance and expressivity of alleles for autogeny in Culex pipiens (L.), Ae. togoi, and Wyeomyia smithii (Coquillett) (Lounibos et al. 1982). Our study demonstrated that larvae raised in the presence of greater food availability emerged as larger adult mosquitoes (as indicated by significantly greater wing lengths) and showed a significant increase in the size of autogenous egg rafts. Laboratory studies had shown that a higher number of autogenously developed ovaries were recorded in Culex annulirostris (Skuse) when raised on a nutrient-rich, compared with a nutrient-poor, diet (Kay et al. 1986). Additionally, Nayar (1968) found a higher number of stage V follicles in Ae. taeniorhynchus raised on a high-quality larval diet.

As wing length has been shown to be a reliable measure of mosquito body size (Nasci 1986), and subsequently a measure of relative nutrition during immature development, positive relationships between wing length and autogenous egg development have been reported (Hugo et al. 2003). In our study, there was a significant relationship between mean wing length of Cx. molestus and the number of eggs in each autogenous egg raft. Linear regression and separate slope analysis demonstrated that the relationship between the number of eggs per autogenous egg raft and wing length size differed based on the different food regimes. Our study also suggested that the size of autogenous egg rafts can be predicted from the wing length size of Cx. molestus in the field. This result has been reported for other Australian autogenous species such as Aedes vigilax (Skuse) (Hugo et al. 2003). However, as noted by Hugo et al. (2003), there are factors other than larval nutrition that may influence the fecundity of autogenous mosquitoes.

Sugar feeding by adults had an influence on the expression of autogeny by Cx. molestus. When both sexes of Cx. molestus were maintained with access to sugar, there was a significant increase in both the quantity of autogenous eggs per raft as well as hatching success. Similar studies indicated that restricting access to sugar resulted in no significant differences for the same parameters with Cx. tarsalis (Su and Mulla 1997). However, our study showed that Cx. molestus can still produce viable autogenous egg rafts when raised on water alone. Further work is required to determine if the significantly smaller-sized autogenous egg rafts and lower rate of successful hatching is due to either female or male nutritional levels, and that a proposed explanation for the result may be reduced fitness of male mosquitoes raised on water alone. It is interesting to consider that Cx. molestus is likely to have limited access to a source of sugar within the subterranean habitats and that, based on the results of our investigation, populations of Cx. molestus could still be maintained under what may otherwise be substandard conditions.

The blood-feeding experiment in our study highlighted an interesting response in Cx. molestus in the presence of a potential blood meal. The result confirmed our earlier work showing that Cx. molestus laid obligatory autogenous egg rafts (Kassim et al. 2011). While other mosquitoes that are obligatory autogenous, such as W. smithii (O'Meara and Lounibos 1981), may mature more than one autogenous egg batch without the requirement of a blood meal, our studies indicated that Cx. molestus does require a blood meal for additional egg batches. However, what our studies also indicated is that unlike some other autogenous species, such as Ae. vigilax (Hugo et al. 2003) and An. amictus hilli (Sweeney and Russell 1973), Cx. molestus will not take a blood meal prior to laying their first batch of autogenous eggs.

We found that the activity of host-seeking and blood feeding started approximately after 48 h of laying autogenous eggs by females and this finding is comparable with the field studies of Miles (1977), which suggested that females of Cx. molestus seek a blood meal only after laying the autogenous egg rafts. Klowden (1990) believed that host-seeking behavior may be inhibited as eggs are developed by the female mosquitoes like Aedes aegypti (L.). Klowden and Lea (1979) found blood feeding influenced mosquito host-seeking in Ae. aegypti, with abdominal distention of females inhibiting their host-seeking behavior, even in the presence of a host. There are also other mosquitoes such as Aedes churchillensis (Ellis and Burst), which has lost flight capability (Hocking 1954), and the arctic species Aedes rempeli (Vockeroth) (Smith and Brust 1970) that never take a blood meal and produce all their eggs autogenously in a single batch.

Several questions arise from these results. While the processes regulating the inhibition of host-seeking behavior in anautogenous mosquitoes have been documented (Brown et al. 1994), the regulation of host-seeking behavior in autogenous mosquitoes prior to oviposition has not been well-studied (O'Meara and Petersen 1985). Perhaps the evolution of this behavior is in response to the adaptation of this species to subterranean habitats where potential blood meals may be scarce. With the exception of rodents, there are unlikely to be available hosts and, by developing autogenous eggs, the mosquito can ensure that the next generation is raised in association with known suitable habitats. Given the cryptic nature of the subterranean habitats, mosquitoes may not be able to relocate their immature habitats if they must leave to find a blood meal. There are many unanswered questions regarding the ecology of this species and the host-seeking and oviposition behavior of Cx. molestus once it has left the subterranean habitats.

Understanding the factors that may influence the expression of autogeny in Cx. molestus can assist in the assessment of this species’ role as a vector of disease-causing pathogens. While the expression of autogeny in Cx. molestus may ensure that this species maintains abundant populations in close proximity to humans, the expression of high rates of autogeny by this species may, in turn, decrease the risk of arbovirus transmission. Fanning et al. (1992), in their investigation of Culex sitiens (Wiedemann), found that the low rate of autogeny of this species may enhance the role of this species as a vector as it maximizes the opportunity to feed on an infective reservoir host and, subsequently, increase the opportunity to transmit disease-causing pathogens to humans. However, even with mosquito species that express autogeny, such as Ae. vigilax or Cx. annulirostris, their willingness to take a blood meal prior to laying autogenous eggs will provide the potential for them to obtain an infected blood meal and become infective sooner. For a species such as Cx. molestus that expresses high rates of autogeny and will not take a blood meal prior to laying their first batch of autogenous eggs, the potential importance of this species for local transmission of disease-causing pathogens may be reduced. Further investigations into the daily survival and overall longevity of this species, as well as vector competence, host preferences, habitat associations, and dispersal, will be required to fully assess the importance of this species to mosquito-borne disease risks in Australia.

Acknowledgments

We thank Merilyn Geary and Karen Willems of the Department of Medical Entomology for their assistance and advice on mosquito rearing and colonization techniques. We thank Karen Byth, statistician with Westmead Hospital (University of Sydney) for providing advice on the analysis of data. We also thank the Malaysian government for supporting Nur Faeza A. Kassim with a national grant for graduate work in Australia.