Theory and application
Sterile insect technique (SIT) has been operational since the late 1950s. Despite the many SIT successes with several insect species during the last four decades, there are still scientists and administrators who do not fully understand the principles or the limitations of the technology. Yet the practitioners of the technology report that as a component of area-wide integrated pest management (AW-IPM) programmes, SIT is currently saving billions of dollars annually in commerce, export markets, reduced environmental impact from pesticides and reduced losses to pests. The initial beneficiaries of this technology were the cattle industry and the wildlife that were spared the ravages of the New World screwworm, now eliminated from North and Central America south through Panama.
Knipling's theories [
1] focus on AW-IPM concepts, under which SIT can target both large and small areas of pest infestation for elimination, suppression or prevention. Depending upon the specific objective and the population characteristics of the target species, SIT can very rarely stand alone and is more likely to be used in combination with prior suppression and/or complementary concurrent control activities. Initiation of AW-IPM programmes, in the USA for example, usually entails governmental and user (shareholder) agreements, which may require formal referenda in which two-thirds of the shareholders must agree. Prior planning for such programmes includes in-depth economic analysis based on capital investment, long-term costs and benefits, and comparison to conventional approaches. Public education is a key component of AW-IPM programmes, which usually include several complementary modes of control. Perhaps the most important aspect of SIT is the realization that it is suitable for only a select group of pests and situations - determined by pest biology, geography, economics and political climate.
Some key misunderstandings have led prominent scientists to underestimate the flexibility and utility of SIT. For example, the target insect need not be monogamous. It is important, however, that sperm of sterilised males be competitive with sperm of the wild males. It is not mandatory that a high ratio of sterile males be attained at the onset, although this certainly would be a desirable situation. The necessary level of over-flooding depends on the biology of the insect, the competitiveness of the released insect and the complementary methods of control that are being considered. It is not absolutely necessary that the infestation be isolated from other sources of the pest, but when isolation is not possible, the programme must have the capability of eliminating the influence of immigrating fertilized females. It is not absolutely necessary to eliminate the target species from an experimental plot to demonstrate that the technique works. Carefully planned research can show what the specific impacts of the releases and other factors, such as immigration, have been. With this information, managers can predict with a high level of probability what can be achieved in operational programmes and avoid the usually impossible task of finding the perfect field plot for proving that the system works.
Perhaps most confusing is the relationship between numerical release requirements and the biotic (reproductive) potential of the pest. Biotic potential usually varies seasonally and may be density dependent. If the effective over-flooding sterile male ratio is 9:1, and the biotic potential (rate of increase R
0) of the target population is five-fold per generation, then the second generation natural population will be only 50% of the initial density (Table
1). However, if the biotic potential is 0.5 (50% decline in population density per generation), the same release ratio will reduce the natural population to 5% of the initial density in the second generation. Rates of increase vary with location, season and environmental conditions. These and other quantifiable parameters are determined during the research and pilot study phases of SIT development, and then further adjusted in the operational phase. If the initial release ratios are sufficient to reduce R
0 to less than 1.0, continued release of the same number of insects will cause population decline. In subsequent generations, the increasing level of induced sterility will eventually lead to elimination of the wild population. Computer models are available for assessing SIT strategies and outcomes [
2,
3]. These simulation models can provide valuable insight for programme planning and conduct, but generally provide the basic framework rather than the specific details for a given pest.
Table 1
Theoretical population trend with nine sterile males released for each initial fertile wild male per generation with a 5-fold rate of increase (after Knipling [
1])
1 | 1,000,000 | 9,000,000 | 9:1 | 500,000 |
2 | 500,000 | 9,000,000 | 18:1 | 131,580 |
3 | 131,580 | 9,000,000 | 68:1 | 9,535 |
4 | 9,535 | 9,000,000 | 944:1 | 50 |
5 | 50 | 9,000,000 | 180,000:1 | 0 |
The SIT approach involves sophisticated technology and experienced staffing is required to effectively manage the rearing processes and distribution of sterile insects. Use of live organisms to control wild populations of the same species can generate unexpected problems. Knowledgeable, well-trained leadership and staff continuity are prerequisites. As an example of what can be accomplished, there are now about 20 fruit fly rearing facilities around the world, one of which currently produces over three billion sterile male Ceratitis capitata pupae weekly for distribution to contracted clients.
Large programmes may continue for decades. The New World screwworm programme took more than 40 years to gradually eliminate the infestation in North and Central American and establish a barrier zone in Panama. Administrators must be convinced about SIT needs and objectives and be dedicated to extended programme continuity and fiscal support. The cost-benefit analysis must justify the long-range commitment.
Ionizing radiation
The first sterile mosquito releases were conducted in 1959-1960 by the United States Department of Agriculture (USDA) in South Florida with males that emerged from radiation-sterilised (120 Gy)
Anopheles quadrimaculatus pupae [
5]. Releases totalled ca 32,000 (over three months) and 300,000 (over nine months) in 1959 and 1960, respectively. Lack of sterility in the wild population led to studies that implicated a changed mating behaviour of the colonised male mosquitoes leading to reduced incidence of mating with wild females and the consequent failure [
6].
Ineffectiveness of released radiation-sterilised (110-180 Gy)
Aedes aegypti in Pensacola, Florida for two mosquito seasons (1960-1961) by the Centers for Disease Control (CDC) [
7] was later attributed to reduced competitiveness caused by irradiation in the pupal stage. Releases totaled 3.9 million in 1960 (four months) and 6.7 million (six months) in 1961.
Effective releases of irradiated (60-120 Gy)
Culex quinquefasciatus were conducted by the World Health Organization/Indian Council of Medical Research (WHO/ICMR) in India and the USDA in Florida between 1967 and 1974 [
8,
9]. These studies confirmed previous laboratory findings that pupal irradiation can be detrimental; but somatic damage was lower with older pupae than with 0-24 h old pupae and competitiveness was reported to be unaffected when one day old adults were irradiated. Releases of sterile males generally ranged from 9,000-15,000 daily in several separate experiments.
In 1980, the first phase of a two year seasonal release study in California,
Culex tarsalis males were determined to be fully competitive following irradiation (60 Gy) as adults [
10]. However, in the 1981 releases, there was a significant inability of the sterile males to seek out and mate with wild females. Lack of control despite adequate over flooding ratios was attributed to assortative mating brought about by selection during colonisation a few months prior to the 1981 releases. A total of 71,000 and 85,000 sterile males were released in 1980 and 1981, respectively.
Chemosterilisation
Several alkylating agents and aziridinyl compounds have been tested as mosquito sterilants. Successful laboratory ([
11], for example) and small-scale field ([
12], for example) studies of these chemicals led to large scale field assessments with
Cx. quinquefasciatus in India by the WHO/ICMR Unit in New Delhi,
Anopheles albimanus in El Salvador by USDA and CDC at the Central America Malaria Research Station in San Salvador, and other species elsewhere [
4].
As part of a broad WHO/ICMR experimental programme on genetic control options for mosquitoes, chemosterilisation studies were initiated with
Cx. quinquefasciatus and
Ae. aegypti in New Delhi around 1971. Thiotepa was selected from several sterilants assessed. Sterile males treated as pupae and released as pupae or adults in several separate field studies were found to be competitive and capable of inducing sterility in wild mosquito populations [
13]. Up to 300,000 sterile males were released daily. However, the study areas were subject to immigration of fertilized females from other breeding sites, whose fertile egg masses impacted severely on the level of sterility and control ultimately achieved. Nevertheless, the suitability of the chemosterilised males was confirmed. The breadth of the investigations touched on many essential aspects of the methodology of conducting control operations with sterile mosquitoes.
In El Salvador, experimental releases of chemosterilised male
An. albimanus were initiated at the beginning of the major breeding season in 1971 at Lake Apastapeque, about 40 km from San Salvador [
14,
15]. Pupae, sterilised by 60 min immersion in an aqueous bisazir chemosterilisation solution [
16], were placed in emergence pans in wooden shelters in natural breeding areas. Starting with an estimated over-flooding ratio of only 2:1, the wild mosquito population was reduced to the point where neither immature stages nor adults could be detected after about five months, suggesting a high level of competitiveness and dispersal ability among the released insects. The rapid increase in vector density normally associated with the malaria transmission season was prevented by these releases. Up to 40,000 sterile males were released daily at the 14-15 km
2 site.
This experiment was followed by a larger scale (150 km
2) pilot study to test the capability of releasing one million sterile males daily in an integrated programme for control of
An. albimanus in a mountainous coastal region [
17]. To minimize the potential for transmission by released females, special attention was given to improved separation of the sexes (separation by size alone had yielded 86% males and 14% females in the Lake Apastapeque trials). A dual sexing system was therefore adopted. Following separation by pupal size and sterilisation, the emerged adults were held in cages for 2-3 days and then offered a bloodmeal containing insecticide (malathion). After the females had fed and died, the males were collected and released. However, these males were not competitive and, furthermore, only ca 40% of the males produced actually survived long enough to be released. Field studies then revealed that sterile males that had emerged from pupae placed in the field were more competitive, whereas males held in cages for one day were less competitive, and males held in cages for 2-3 days were unable to induce sufficient sterility into the indigenous population.
To attain maximum production and competitiveness, the colony strain was replaced by a strain with a chromosome translocation linking propoxur resistance to the Y-chromosome [
18]. Exposure of eggs to propoxur killed the females but not the males. Twice as many males could then be produced from each rearing unit and released in the pupal stage. Males of this genetic sexing strain (MACHO) performed very well in the field, reducing the wild population in a small (20 km
2) experimental block to a fraction of its prior density while the population in the untreated control zone increased several fold. The releases were then integrated with larval control measures and the population was further reduced to one-tenth of its original density (97% control in four months). At this stage of the research, the special support funds had been exhausted and El Salvador was in a state of civil disruption. Studies were discontinued.