Species complexes are common in the class insecta, (Subbarao 1997). Where morphologically identical populations do not interbreed separate biological species may exist, these are known as sibling, cryptic or isomorphic species. There are many examples of such complexes most notably the Anopheles gambiae complex, the most important malaria vector world-wide. This paper will concentrate on the Indian malaria vector Anopheles culicifacies Giles sensu lato (Diptera: Culicidae), and the techniques available for the identification of its sibling species and their importance in malaria incidence. Anopheles culicifacies is the major vector of malaria in India, and is responsible for the transmission of 60-70 per cent of the 2-3 million malaria cases reported annually, (Subbarao 1997). This taxon exists as an isomorphic species complex consisting of four reproductively isolated populations. Green and Miles (1980) first discovered two populations, designated A and B, from studying the ovarian polytene X-chromosomes in half-gravid females. A further two species were later described and designated species C (Subbarao 1983), and D (Vasantha 1991), all four species are distinguishable by paracentric inversions on the X-chromosome. Polytene chromosomes provide accurate identification, however there are certain disadvantages. Identification is limited to semi-gravid blood fed females, the technique requires a good deal of skill to perform, and the number of specimens that can be processed in a given time is also limited, (Phillips 1994). A great deal of work has been done to develop an alternative method that is simple, effective and available to the non-molecular biologist in the field. Therefore various techniques are available for An culicifacies sibling species identification, these are structural variations in the Y-chromosome, cuticular hydrocarbon profiles, lactate dehydrogenase alloenzyme differentiation and DNA probes.

Differences in the structure of the male mitotic Y-chromosome were first reported by Vasantha ,1982 & 83, who noted differences between species A, B and C. A and B appeared to have sub-metacentric Y-chromosomes whereas B was acrocentric, enabling the identification of B from A and C. This approach involves the use of chromosomes in the larvae, therefore semi-gravid females aren't required. However, if specimens are taken from indoor resting sites, which is important for epidemiological studies, they have to be reared for F1 larvae examination, this is very impractical and time consuming. Furthermore a recent study has demonstrated that in fact the metacentric polymorphism between the three species is not as rigid as first thought, species B and C do exist in acrocentric and sub-metacentric forms, therefore differentiation of the members of the complex based on male karyotypes is inaccurate, (Adak 1997). The technical skill required and time needed for chromosomal identification promoted a search for alternative techniques. Hydrocarbon profiles involve the analysis of cuticular components through gas liquid chromatography (GLC). The profiles of each isomorphic species can be identified from the varying retention times of the cuticular hydrocarbons. Milligan, 1986, suggested that the three cytospecies, A,B and C, were found to be significantly different in their cuticular hydrocarbon composition. Further this technique has many advantages over that of polytene chromosome identification. The material could be stored and analysed in any condition and both sexes and all adult stages can be identified, which is impossible with chromosomal examination. However, this approach does involve expertise, sophisticated equipment and more importantly is very time consuming, with GLC runs taking about half an hour for each specimen, ( Priestman, personal communication). Allozymes are different molecular forms of an enzyme coded by different allelic forms of a gene, therefore after electrophoresis the enzymes activity can be visualised on a gel and the species identified. Adak, 1994, isolated an allozyme, lactate dehydrogenase, which could be used to distinguish species A and D from species B and C in the complex. The expression of lactate dehydrogenase is autosomal therefore all life-stages can be identified, moreover it is a genetic expression therefore variation between individuals and populations would be minimal and unaffected by environmental factors, which is a criticism of hydrocarbon profiles, (Phillips 1988). Although similar advantages to hydrocarbon analysis exist, this technique is time consuming and relatively expensive in terms of equipment and materials, (Hill 1994).

The most recent advancement in An. culicifacies complex identification has been the development of DNA probes. Probes are specific stretches of RNA or DNA that are labelled and can hybridise with complimentary nucleic acid stretches. DNA Probes have now been developed that can distinguish species A from species B and C, (Gunasekera 1997). Any mosquito, stage or part can be used therefore specimens can be stored and identified at a later date with other probes or future techniques. The process is quick, and there are few problems for storage or handling enabling testing of frozen dried or alcohol preserved specimens, (Hill 1994). The current probes that have been developed are radioactively labelled which would provide problems of storage, handling and disposal, (Gunasekera 1997). However, if radioisotopes for labelling and synthetic oligonucleotide probes can be developed, as has been for the identification of the An gambiae complex, there is potential for a safe, quick, easy and reliable identification method that could be used in the field, (Hill 1994) Due to limitations of methods using biochemical identification and DNA probes the diagnostic inversion of polytene chromosomes remains the most accurate method of identification of An. culicifacies sibling species, (Adak 1997). Although due to recent advancements in DNA and isoenzyme techniques this method may soon be replaced. The literature on the An culicifacies complex has applied the polytene chromosome identification technique, due to its reliability, to species identification. Different species within a complex may exhibit differences in ecology, vectorial capacity and response to control measures, (White 1982). Many examples exist that highlight the importance of the identification of sibling species in the An. culicifacies complex. The initial observation, made by Green and Miles in 1980, of the two sibling species A and B has had many ramifications in malaria control strategies. The most important difference between the two species is in their susceptibility to Plasmodium. Today it is generally accepted that species A is the primary vector of both Plasmodium falciparum and P. vivax and that species B is a non-vector species, (Subbarao 1988a). In Western Uttar Pradesh, India, it was demonstrated that where species A and B were found sympatrically with a predominance of species A, both P. falciparum and P. vivax were present and the malaria incidence remained high. Whereas in three eastern districts that had a predominance of species B with an occasional occurrence of species A malaria incidence was almost absent, (Subbarao 1988b). Species C and D have also been incriminated in malaria transmission with C more so than D however both are less important than A, (Subbarao 1992).

Susceptibility to Plasmodium is one of many parameters important in vectorial capacity. The level of anthropophily has been observed using immunoelectrophoresis to assay bloodmeals. Joshi, 1988, indicated that although species A and B are predominantly zoophagic A is more anthropophilic than B. Further the HBI was found to be related to the proportion of human and cattle population in the area, suggesting that increasing the cattle population may reduce transmission.

Finally, an extremely important distinction that can be made between the four sibling species is their susceptibility to insecticides. Various papers have described contrasting levels of susceptibility to insecticides throughout India. Subbarao, 1988c, reported resistance to DDT, species B was found to be less susceptible than A. Further both species demonstrated resistance to HCH due to the long programmes of house spraying. Raghavendra, 1991, 1992, reported varying levels of resistance to malathion, where species B and C are sympatric B is more susceptible than C, also with species A and B species A is more susceptible than B. Therefore in control programmes the type of insecticide is extremely important. For example, if A and B exist sympatrically and DDT and malathion was sprayed the population sensu lato may not change but transmission rates will decrease, due to a shift in prevalence from A to B. This also highlights the importance of sibling species identification in the monitoring of control strategies. Differences between sibling species which are important in relation to vectorial capacity, or insecticide resistance can be used to help predict the most effective approach for control programmes. Historically the main method of vector control was widespread house spraying, however species identification has enabled the targeting of strategies, such as house spraying, and the promotion of other strategies such as Bio-environmental control and bed nets. This is important as the high costs of insecticides especially in areas with emerging resistance makes spraying economically non-viable as along term control strategy, (Subbarao 1994). With the aid of reliable, easy to use field based identification techniques existing control strategies could be optimised consequently reducing malaria transmission. Subbarao 1997, has applied the knowledge of the biological differences among sibling species to malaria control in Uttar Pradesh. For example in the east, where species B is dominant and transmission is low, chemotherapy on imported cases would be appropriate instead of the widespread application of insecticides. In Central and Western areas, that are prone to epidemics, selective seasonal spraying during peak times combined with bio-environmental control during inter epidemic periods would be most cost effective. If this stratified approach could be applied at the highest resolution, made possible with the use rapid identification techniques, pockets of efficient vectors could be isolated and controlled at the ultimate efficiency both reducing side effects and optimising costs.