Crowding of Culex sitiens Wiedemann larvae was investigated to determine whether pupation was delayed by the presence of chemical factors in the water or by mechanical interference. As the larval density increased from 0.05 to 1.0 larvae per milliliter of water, the duration of larval development increased from 7.5 to 23.0 d and larvae mortality increased from 1.7 to 99.6%. At high larval densities (2 larvae per milliliter), the larvae did not pupate, but remained alive as 4th instars for at least 50 d, even after their abundance subsequently declined due to mortality. When not crowded, 80% of the larvae fed at the surface; the remainder fed mainly on the bottom (even at a depth of 35 cm). Removal of chemical retardants by changing the water daily increased pupation to 65% (compared with 4% in controls, in which the water was not changed), but this was less than the effect of reducing mechanical interference by increasing the surface area, which increased pupation to 98%. Therefore, the main density-dependent factor was mechanical interference during feeding, but prolonged development to 19 d showed that chemical retardants also were important. The chemical retardant was not stable, so that 2nd generation larvae reared in the same water were unaffected, but its effect was irreversible, because most crowded larvae moved to clean water and a lower density did not pupate.

Culex quinquefasciatus Say (Diptera: Culicidae) is an abundant urban mosquito that is the vector of filariasis. Breeding in septic tanks, where there are very high levels of bacterial food, it is likely to have a different reaction to crowding compared with other mosquitoes. To test for the presence and type of crowding effects, four larval densities of C. quinquefasciatus varying from 0.4 to 3.2 larvae ml−1 of water were reared in tubes. Mortality was found to greatly increase at densities above 0.8 larvae, whereas larval duration increased even above 0.4 larvae ml−1. Changing the water in the tubes daily gave a small (but significant) response in reducing mortality and larval duration. However, when larvae kept at a low density shared the same water with larvae at high density, there was no chemical influence on their growth rate and mortality. The effect of crowding was primarily due to physical disturbances between larvae. When larvae were kept at a high density in the same volume of water, but in shallow trays with a large surface area and therefore much less contact between them, mortality was the same as for the lowest density. There was still, however, a significant increase in larval duration from 8.6 days at 0.4 larvae ml−1 to 12.1 days at 3.2 larvae ml−1. It is therefore concluded that the larvae respond to physical rather than chemical factors by prolonging larval development and having some increase in mortality.

No abstract

Four species of mosquitoes were abundant in concrete reservoir tanks containing brackish water that ranged from 16 to 39% sea water. The ability of the larvae to survive in various salinities was compared for each species with the ovipositional preferences of the adult females to determine whether the 2 traits were correlated. Southern house mosquito, Culex quinquefasciatus Say, normally was not present in the tanks but survived well in salinities up to 25% sea water. However, gravid females almost always oviposited in fresh water. Culex sitiens Weidemann larvae survived best in saline water (66% sea water), but oviposition was greatest in 28% sea water; both larval survival and the frequency of oviposition were low in fresh water. Culex sinaiticus Kirkpatrick survived salinities up to 50% sea water, but the females refused to blood-feed; therefore, their ovipositional preferences were not tested. Larvae of Anopheles stephensi Liston and An. culicifacies Giles survived best in fresh water, but some An. stephensi were able to tolerate up to 50% sea water. The females had a similar ovipositional preference for fresh water. The preferred salinity for oviposition did not correspond with larval survival for Cx. quinquefasciatus and Cx. sitiens, but did compare well in An. stephensi and An. culicifacies.

Although predators have been extensively used in the biological control of mosquito larvae, their efficacy will be reduced if the larvae are able to detect and respond to their presence. This ability to detect and respond to the unseen presence of dragonfly or damselfly larvae was investigated for the larvae of three mosquito species (all Diptera: Culicidae) to see whether they either altered their rate of development or their feeding behaviour. The development of Culex quinquefasciatus Say and Culiseta longiareolata Macquart larvae was not affected by exudates from either predator, but Culex sinaiticus Kirkpatrick developed significantly slower when reared in the presence of a caged dragonfly larva and produced adults that were significantly smaller, indicating that they probably reduced feeding activity to make themselves less detectable to the predator. This only occurred when the dragonflies were fed in situ (in the presence of the mosquito larva) and not when removed for feeding, so that although this removed predator kairomones, other cues such as vibrations caused by movement of the predator may also be involved. In addition, the mosquito larvae responded to dragonfly larvae, but not to damselfly larvae. The depth of the water in which the mosquito larvae were reared had no effect on their response to the dragonfly. All three mosquito species significantly increased surface filter‐feeding, when a predator was present in the water (compared with controls where no predator was present), at the expense of bottom scraping. In Cx. sinaiticus (but not the other two species), surface filter‐feeding was greater when a dragonfly was present, compared with a damselfly. In the experimental conditions of one predator per 250 ml water, all three mosquito species were thus able to detect dragonfly and damselfly larvae (and distinguish between the two), but their response varied among species.