These insects are a major pest species causing myasis in both livestock and man. This report will look at damage caused by these flies and the use of sterile male techniques to control them.

Cochliomyia hominivorax is an obligate parasite of endothermic vertebrates, whilst Cochliomyia macellaria is a facultative parasite, with carrion being the preferred ovipostion site.⁽¹⁾ Both species are referred to as screwworm and the invasion of body tissues by their larvae is termed myiasis. The economic loss due to myiasis from screwworm was estimated at $650m annually in 1985, with 90% of myiasis cases examined due to C. hominivorax⁽²⁾ hence the common name of primary screwworm. Until 1933 the two species were considered to be conspecific but morphological features of the male genitalia allowed the two species to be distinguished.⁽³⁾ The control of C. hominivorax using the sterile insect technique (SIT) has been extremely successful in North America and Mexico. Despite the prolonged release of sterile males however there are still sporadic outbreaks in the south-western United States and northern Mexico. An explanation for these outbreaks is the possibility of premating isolating mechanisms within screwworm populations. The implications of such isolating mechanisms are that sterile males may be ineffective due to non-mating.⁽⁴⁾ The ensuing essay will endeavour to explain genetic flow in screwworm populations in relation to both the flies biology and the SIT control programme.

Before embarking on the details of gene flow and the evidence for a panmictic population, a brief description of the ecology of the two species will help to explain the population genetics. The two species are considerably divergent in both their density in the field and the larval environment that they exploit. C. hominivorax is an obligate parasite and a female will only oviposit on a wound site, such as that inflicted by a feeding tick. A female will lay around 300 hundred eggs which will hatch after 11-21 hours, the larvae eat into the flesh and in four to eight days leave the host to pupate in the soil. Above 160C C. hominivorax will be active all year, overwintering usually occurred in Texas and in the summer wide dispersal from this region resulted. Female C. hominivorax have been shown to disperse 56km a week,⁽⁵⁾ with a maximum dispersal measured at nearly 300km.⁽⁶⁾ The range of C. macellaria is much wider both geographically and in larval environment. C. macellaria is found from southern Canada to Chile and Argentina, with wound sites on vertebrates being second choice oviposition sites to carrion (see footnote). This ecology gives rise to the name secondary screwworm. In 1936 the ratio of C. hominivorax to C. macellaria within the field at wound sites was 1: 590⁽⁷⁾ showing how large C. macellaria populations are in relation to C. hominivorax. This differential would be even greater today due to the successful control programme. This does not reflect that C. macellaria is more important as a myiasis agent as it often feeds at wound sites but does not oviposit.

The implication of these population structures is that heterozygosity of C. macellaria will be larger than that seen for C. hominivorax, and the FIS⁽⁸⁾ values will be lower. This is due to higher genetic variability in species with more diverse habitats and the high population density makes it more unlikely that samples used for testing will be siblings. Allelic differences are illustrated in the table 1 below.

Considering that the control programme implemented against screwworm involved a genetic technique it would seem reasonable to assume that there would be a plethora of papers on the genetic structure of screwworm populations. This is not the case however. One of the early papers that looked at genetic structure of screwworm populations back in the early seventies studied enzyme loci and identified five variable regions within C. hominivorax. This studied only looked at laboratory colonies though and so was not truly indicative of allozyme polymorphism in field populations⁽⁹⁾ and thus did not reveal the extent of gene flow in screwworm populations. In the late 1970’s and early 1980’s other studies, one using cytological and isoenzyme data reported that the screwworm population in Texas, Mexico, and south of Mexico, harboured many cryptic species of screwworm many of which were sympatric.^(4,10) Richardson and colleagues proposed the existence of many isolated populations of C. hominivorax based on allozyme polymorphism(see footnote). Although the population structure is very fragmented across its range migration and widespread dispersal is part of the fundamental biology seen in this species. It should not be construed from this migration and dispersal however that a focal population will radiate out analogous to ripples emanating from a stone thrown into water. Female C. hominivorax are very fecund and only a small number need to migrate, this will be adequate to cause the rapid colonisation of a new area.

Indeed if interbreeding does not occur in screwworms selection for certain alleles occurs very quickly. This will be discussed later when looking at the SIT control programme in relation to fitness between wild and reared males.

The studies by Richardson et al., cannot be replicated to substantiate the evidence for cryptic species due to the success of the eradication programme. It could be construed that the success of the control programme is evidence against cryptic species as populations of cryptic species would not mate with sterile released males, due to some phenotypic manifestation i.e. contact pheromones within some species of screwworm.

Fig 1 Map showing six sites where C. hominivorax were collected. MN48 Michoacan, Mexico. CIH34 Colima, Mexico. 009 Texas. JAM3b Jamaica. EG2 Oaxaca, Mexico. SP94 Chiapas Mexico. See text for explanation.

An interesting study using mitochondrial DNA from C. hominivorax collected from six geographic regions throughout the range of this species showed less diversity in mainland C. hominivorax compared to C. hominivorax on Jamaica.(11) When pairwise comparisons were made including Jamaican C. hominivorax the percentage of sequence divergence’s were all > 2.1%. When pair combinations were made between even the most widely separated mainland samples the percentage sequence divergence was < 0.7%.

The percentages of all comparisons can be seen in figure 1 above. This study indicates some isolation of screwworm on Jamaica but little on the mainland. It is reasonable to assume that flies are less likely to cross a wide water barrier than cross areas of land. The main route across the sea would be via livestock, this is mostly one way however, from the US to Jamaica and therefore there is little chance of introducing flies by this route.

Other workers have calculated the F statistics for C. hominivorax populations from Costa Rica and Mexico.⁽¹²⁾ Three loci were studied and resulted in FIS = 0.11, FST = 0.01 and FIT = 0.12. The values for FIS and FIT are relatively large compared to the FST. These values are due to the Wahlund effect which arose due to the pooling of data from more than one paper and splitting it into three groups. The FST scores indicates low genetic diversity and this can result from only small amounts of migration between these populations. Nei’s genetic distances between Costa Rica and Mexican populations were 0.005 - 0.008, showing close similarity in both C. hominivorax and C. macellaria populations.

Having discussed studies by several workers in this field it is worth focusing on the findings of these workers and how the knowledge gleaned from these studies help the SIT programme. The need to elucidate the genetic structure of C. hominivorax and C. macellaria populations is fundamental in understanding how best to implement a control programme. The economic impact of C. macellaria is only a fraction of that caused by C. hominivorax. It is controllable to a certain extent with the use of topical insecticides. The high density of C. macellaria across such a vast area would not make a SIT campaign commercially viable or achievable as a ratio of 10:1 sterile too fertile males is optimal.

The SIT programme against C. hominivorax has been a huge success and the screwworm population is being driven back into an every decreasing range. The existence of cryptic species within C. hominivorax would seem feasible due to their aggregated population formation throughout their range. It is their high propensity for migration that removes the possibility of isolated breeding populations such as that proposed by Richardson et al.(See footnote). Therefore when combined with the published literature which is nearly all orientated towards a single species concept resulting in a large panmictic population the likelihood of sympatric cryptic species seems impossible.

An analogous situation could be proposed with the Anopheles gambiae complex in Africa. The morphologically identical An. gambiae and An. arabensis have been distinguished by various PCR methods.⁽¹³⁾, and in the field mating between the two species are extremely rare. A SIT campaign using either Anopheles spp. would fail to give an eradication similar to that achieved against C.hominivorax.

A genetic problem did arise during the SIT campaign when selection for non-competitive ecotypes occurred in the rearing procedure.⁽¹⁴⁾ The enzyme a-glycerol phosphate dehydrogenase (a GDH) was discovered to have two allelic forms. Selection for one of these forms resulted in a change in male activity in the field. Wild males are active from early morning, whilst sterile males would not become active until the afternoon, thus wild males had time to mate without the competition from sterile males.

To conclude, there is unrestricted gene flow in populations of both species and genetic differentiation that arises by drift or selection is homogenized by migration resulting in panmictic populations, this is supported by the Hardy-Weinberg values in table 1.