Eradication of the cattle tick from the USA

In 14 of the southern states of the USA, a campaign was started in 1906 to eradicate redwater fever from cattle. By 1943 it was almost complete, with several foci remaining in Florida until 1961. The means of eradicating the causative Babesia bovis and Babesia bigemina protozoa, long before there were any anti-babesial vaccines or drugs, was by eradicating the vector ticks, Rhipicephalus (Boophilus) annulatus and Rhipicephalus (Boophilus) microplus. (The former genus Boophilus is now a sub-genus within Rhipicephalus, but the above two species, and R. (B.) decoloratus are in a group distinctive for control so in this review are called collectively boophilids where necessary.)

Cattle owners in the north noticed that when southern cattle were trekked northwards, the local cattle along the route contracted soon after a disease with the sign of haemoglobinuria. In 1795 North Carolina passed a law restricting movement of cattle within the state in response to such an undiagnosed new disease. More states followed with similar restrictions of movement of cattle from the south. In this way, pragmatic public health containment of this disease was started without formal knowledge of its causes. Curtice (Reference Curtice1891) first described the obscure life cycle of R. annulatus as a 1-host type. (That is, the larvae, nymphs and adults all feed in a single sequence on one individual host with the moults occurring on that host and only ovipositing females, eggs and questing larvae occurring on vegetation. All boophilid ticks have a 1-host life cycle.) Curtice used his knowledge of generation times and phenology to devise a systematic treatment of cattle. By 1889 the US Department of Agriculture enforced movement restrictions during the summer on cattle that would normally be moved northward for sale. These southern cattle were of Criollo stock, long adapted by exposure to Babesia and R. annulatus, so their owners were untroubled with costs of clinical redwater. Cattle farmers further north, however, suffered losses in their typically European breeds of cattle that were susceptible as adults to fulminating infections with Babesia (Graham and Hourrigan, Reference Graham and Hourrigan1977; George, Reference George1989a). Local governments supported tick-control schemes and set up Livestock Sanitary Boards. These employed field inspectors and built dip tanks charged with crude petroleum oil suspended in water. This type of dip was invented in the 1890s with much development work from Texas A & M University, based on experience with various crude pesticides used for many decades against mites and flies on livestock and horses.

One source of the science of veterinary entomology was from Smith and Kilborne (Reference Smith and Kilborne1893) who, as employees of the Bureau of Animal Industries, discovered how B. bigemina is transmitted between cattle through R. annulatus. Knowledge of the aetiology of the disease, and of its cost of approximately $US 1 billion per year at today's prices, impelled a declaration by the US Congress in 1906 to eradicate the tick down to the border with Mexico. The biological borders would be 2 oceans and northern cold. Methods combined surveillance for infestation, isolation, transport restrictions, dipping mandated by new state laws and inspection for compliance and effectiveness of dipping. Pasture management against ticks was used at farmer's discretion, with rotations of herds between pastures and rotations of pastures and crops (Teel et al. Reference Teel, Marin, Grant and Stuth1997). The acaricidal treatments every 2 weeks for all cattle were the mainstay of the campaign, more so when arsenical washes were introduced in the 1910s. These remained in use until the invention of pesticidal DDT in the mid-1940s.

This eradication was a magnificent achievement with permanent benefits to many. This is despite the cost and difficulties of maintaining a barrier against ticks with multiple acaricide resistances from Mexico. There are, however, two sides to this story. One is told by scientists such as Graham and Hourrigan (Reference Graham and Hourrigan1977) and George (Reference George1989a) who described dedicated veterinarians and biologists collaborating with determined cattle ranchers for the onerous collective endeavour. The other story is told by a social historian (Strom, Reference Strom2009). She described how some small-holding mixed farmers in these southern states expressed fierce opposition to the campaign. They saw costs without profits in the troublesome task of mustering their semi-feral and naturally immune cattle, from wooded and swampy terrain every 2 weeks, for the expensive and noxious chore of compulsory dipping. Dynamite was readily available and a popular form of protest was to detonate it in dip tanks (Strom, Reference Strom2004).

Attempted eradication of cattle tick from New South Wales

The example from America inspired cattle ranchers, and researchers employed by state governments in Australia, to follow. Rhipicephalus microplus and its babesias entered the Northern Territory of Australia from Timor or Bali about the 1830s to 40s, then spread along the moist subtropics to eastern Queensland and further south into New South Wales by 1906 (Angus, Reference Angus1996, Reference Angus2003). In 1920 the Board of Tick Control of NSW recommended a 3-year campaign with compulsory dipping of all cattle in the most heavily infested areas and isolation of individual infested farms in other areas in the range of the tick. The campaign borders were defined by the sea, dry inland and cold southern regions and a fenced barrier zone against Queensland. The campaign struggled from partial successes to reluctant admission of failure by the early 1980s. All the elements of the American campaign were used, except for sufficient compulsion. Opposition to the campaign by farmers was less strident than in America, but with a combination of neither sanctions nor private benefit for many farmers, and the increasing availability of drugs and vaccines against Babesia, failure against this highly invasive pest was likely (Angus, Reference Angus2003).

Eradication of the agent of East Coast fever from southern Africa by tick control

Arnold Theiler, Stewart Stockman and others in South Africa adapted, from 1902–4, methods from the USA to stop their devastating losses of cattle from East Coast fever (ECf). This disease emerged from eastern Africa into a country already suffering catastrophe from the rinderpest pandemic, at a time when draught oxen powered most heavy transports. The infection came in cattle, via Dar-es-Salaam into Mutare in Zimbabwe, probably by sea and railway from Beira. Local veterinarians and consultants such as Robert Koch thought that the new disease was a virulent form of babesiosis. By 1903, however, Koch formally named this as a previously unrecognized disease, caused by Theileria parva protozoa.

When crude vaccines failed, the decision was taken to eradicate the disease, mainly by killing the newly discovered vector – Rhipicephalus appendiculatus. Based on local experiments with pasture rotation, there was much reliance on this method, despite a period of 15 months needed for this long-lived species of 3-host tick. (A 3-host tick has larvae, nymphs and adults each feeding separately on one individual host with both moulting and questing occurring on the vegetation.) In addition, herds suspected of being infected were isolated in temperature camps. Here the courses of individual infections were monitored by daily measurements of temperature. During a conference in Pretoria in 1929, a formal eradication campaign was devised. The central component was intensive surveillance of cattle in temperature camps for 25 days using diagnosis by microscopical examination of blood smears, together with prolonged isolation or slaughter of infected cattle to eliminate foci of infection with T. parva. Mandatory dipping at a maximum interval of 5 days was also very important, and supplementary measures were fencing and rotation of pastures (Lawrence, Reference Lawrence, Norval, Perry and Young1992).

By breaking its cycle of transmission, T. parva was eradicated from the still surviving populations of R. appendiculatus in Zimbabwe by 1954; several years later in South Africa, and the last focus in southern Africa was eradicated from Swaziland in 1960. Thus, it took 30 years to complete this campaign throughout that region (Lawrence, Reference Lawrence, Norval, Perry and Young1992; Norval et al. Reference Norval, Perry and Young1992). This heroic era of large-scale and effective veterinary disease control is recounted in the books by Gutsche (Reference Gutsche1979) and Cranefield (Reference Cranefield1991). In South Africa the natural boundaries were the eastern sea, dry interior and cold in this southerly projected distribution of R. appendiculatus in Africa. The challenge would prove far greater northwards, in central and eastern Africa, where eradications were attempted in regions of increasing climatic favourability for R. appendiculatus, and in countries contiguous with others where the tick remained uncontrolled. Such campaigns also faced the problem of political instability, as discussed below in the section on changes in land use and social factors.

Eradication campaign against the bont tick in the Caribbean

Amblyomma variegatum was eradicated in the early 2000s from 4 islands of the Caribbean in a campaign run by the Food and Agriculture Organisation, in association with the US Department of Agriculture and other agencies (Pegram, Reference Pegram2010). The original promotion for the campaign was to eradicate this tick from the whole region and the implementation phase lasted 12 years. Amblyomma variegatum is not only a vector of Cowdria ruminantium bacteria causing heartwater in cattle, sheep and goats, it also aggravates severely the skin disease dermatophilosis by a systemic immune suppression associated with its feeding (Walker and Lloyd, 1993). This tick arrived on cattle from Africa in about 1830 (Uilenberg et al. Reference Uilenberg, Barré, Camus, Burridge and Garris1984) and it is considered a severe potential threat, as vector of Cowdria, to the cattle industry of the Americas (Bram and George, Reference Bram and George2000; Bram et al. Reference Bram, George, Reichard and Tabachnick2002). In those islands of the Caribbean where local breeds of cattle have long ago acquired the ability to co-exist adequately with both the tick and Cowdria, farmers often regarded these diseases as an insufficient reason to participate with the campaign. Moreover, in the islands cleared of the tick, the effort had to be intense because of the residual populations on large numbers of alternative hosts such as feral donkeys, and stray small ruminants that graze in derelict cash-crop areas (Pegram and Eddy, Reference Pegram and Eddy2002; Ahoussou et al. Reference Ahoussou, Lancelot, Sanford, Porphyre, Bartlette-Powell, Compton, Henry, Maitland, Lloyd, Mattioli, Chavernac, Stachurski, Martinez, Meyer, Vachiery, Pegram and Lefrancois2010). This campaign included 18 nation states or islands and was funded and directed from far away by 5 main international development agencies, but without clear managerial lines of authority. This led to severe operational problems that often confounded the stoical efforts of the small group of scientific and administrative staff based in the Caribbean (Pegram, Reference Pegram2010). As Pegram et al. (Reference Pegram, Wilson and Hansen2000) emphasized from case studies of eradication campaigns, the successes have been associated with an appropriate combination of enforcement and good management.

Natural limits on tick populations

The central question of whether or not ticks have stable populations has been little studied. It is essential to know this for better application of understanding of tick population dynamics to tick control. Ticks have relatively high reproductive potential: A. variegatum females can weigh up to 5 g when engorged and lay more than 20 000 eggs. In a stable population, what kills all but 2 of the ticks that could develop from such an egg batch? Ticks have a reproductive capacity that can be categorized as intermediate between the extremes of high reproduction with low survival of offspring, contrasted to low reproduction with high survival of offspring. Sutherst et al. (Reference Sutherst, Wharton, Cook, Sutherland and Bourne1979b) reported 100-fold fluctuations of R. microplus over 6 years, which they called large differences, but the comparison needed here is with organisms like plague locusts. Furthermore, splendidly long studies have been made on populations of A. hebraeum, R. decoloratus and R. appendiculatus, maintained by wild hosts in South Africa, and sampling by standardized monthly cloth drags for larvae. Over 164 months these populations remained within a 100-fold range of mean numbers collected each year (Horak et al. Reference Horak, Gallivan and Spickett2011; Spickett et al. Reference Spickett, Gallivan and Horak2011). The crucial influence of density-dependent mortality on such stability is discussed below.

Many of these ticks, especially larvae, probably perish from desiccation whilst questing, and birds and other predators consume all stages. What proportion of the population succumbs these ways remains uncertain. Similarly the relative contribution of acquired immunity of hosts against the feeding of ticks is uncertain. Under conditions of experimental infestations, host immunity can be fatal to many of the ticks (Hewetson, Reference Hewetson1972; Jongejan et al. Reference Jongejan, Pegram, Zivkovic, Hensen, Mwase, Thielemans, Cosse, Niewold, Elsaid and Uilenberg1989; Latif et al. Reference Latif, Nokoe, Punyua and Capstick1991a, Reference Latif, Punyua, Capstick and Newsonb). These defensive immune reactions act across the instars, and there is high mortality of larvae of R. appendiculatus attempting to feed on cattle that have acquired resistance by exposure to adults only (Chiera et al. Reference Chiera, Newson and Cunningham1985; Walker et al. Reference Walker, Fletcher and Todd1990).

The ability of individual cattle to acquire immunity to ticks is strongly heritable (Seifert, Reference Seifert1971). There are major differences in this ability between breeds, mainly contrasting cattle of Bos taurus stock which have moderate ability to resist ticks (85% of R. microplus applied will be killed by immune defences) compared to Bos indicus stock with a high ability (98% of ticks killed). Thus taurus Friesians can acquire immunity to ticks, but indicus Sahiwals acquire it about 7-fold stronger in terms of reduction in successful completion of individual tick life cycles. There are behavioural differences between these types of cattle that also reduce tick feeding: indicus cattle are better able to avoid the visible clumps of R. microplus larvae on grass (Sutherst et al. Reference Sutherst, Floyd, Bourne and Dallwitz1986). The exploitation of host immunity for tick control is discussed below, but first its effects on the natural regulation of tick populations need to be considered.

The question of density-dependent mortality in tick populations has been difficult to answer. Nevertheless, the evidence is strong. Experimental studies with artificial infestations of cattle with R. microplus showed a distinct increase in the ability of cattle to mount defence against ticks with increasing levels of infestation by ticks (Sutherst et al. Reference Sutherst, Wagland and Roberts1978). The explanation was that the more ticks attempting to feed, the greater the immunological challenge and thus the stronger the immunological response. A study on captive wild rodents (Clethrionomys glareolus) experimentally infested with one of their natural parasites, Ixodes trianguliceps, showed the same effect, with a similar explanation (Randolph, Reference Randolph1994b). For R. appendiculatus it is regrettable that the large number of experimental studies of infestations of cattle during the 1970s and 80s lacked the resources to tackle this question experimentally. However, statistical re-analysis of the results of various studies on natural infestations of cattle by this tick revealed an effect from stage to stage of the ticks that demonstrates strong density-dependent mortality (Randolph, Reference Randolph1997).

Examination of individual livestock animals on farms usually reveals some individuals heavily infested whilst most individuals in the herd have few ticks. This has been commented on (Latif et al. 1991b) but may be less pronounced (Kaiser et al. Reference Kaiser, Sutherst and Bourne1982) than found with other types of parasites, and Kaiser et al. (Reference Kaiser, Sutherst and Bourne1991) found 50% of the R. appendiculatus engorged females on 30% of the herd of cattle. These aggregated, or overdispersed, distributions have been found in a wide range of organisms (Taylor, Reference Taylor1961) and can readily be described by various mathematical distributions, typically power laws. On natural wild hosts aggregation is pronounced (Randolph et al. Reference Randolph, Miklisova, Lysy, Rogers and Labuda1999; Kiffner et al. Reference Kiffner, Lodige, Alings, Vor and Ruhe2010) and may be caused by variation in host immunity or heterogeneities in distribution of ticks on vegetation (Shaw et al. Reference Shaw, Grenfell and Dobson1998). Individual cattle with a genetic poor ability to acquire resistance to ticks are likely to have higher densities of infestation. Suppression of immunity of cattle caused by dense infestations of R. appendiculatus has been reported by Fivaz (Reference Fivaz1989). Thus it is suggested in this review that such immune suppression provides a positive feedback that partially drives aggregated distribution of ticks on their hosts, and that the subject deserves more study.

Timing of treatments in relation to tick life cycles

From the start of action against R. annulatus in the USA there was a problem with the treatment interval. Once every 21 days should break the 1-host life cycle that takes 18 to 22 days. The use of treatment once every 14 days was because of the short period of activity of the early acaricides remaining as a residue on the hair coat of cattle, unlike the lipophilic organochlorines that eventually were used. Without a sufficient residual active period any ticks feeding rapidly during hot seasons might escape treatment every 21 days and thus infest pastures with larvae infected transovarially with Babesia. Thus the burden of treatment for the farmers was approximately one third greater each year of the main campaign compared to the typical 3-weekly treatment for the control of boophilid ticks.

In southern Africa the problem with R. appendiculatus and T. parva was more complex, as Charles Lounsbury found amongst the confusion of theileriosis with babesiosis. Veterinary staff investigating outbreaks observed the small-scale patchy distribution of new cases within a herd, leading them to suspect ticks as the vector. Lounsbury (Reference Lounsbury1903) in his experiments with ticks, assumed transovarial transmission, by extrapolation from knowledge of Babesia transmission. When his collaborators supplied him with some nymphs and adults of an unfamiliar species of tick he then realized that his knowledge of transmission of B. bigemina was insufficient to explain theileriosis. Rhipicephalus appendiculatus had been formally described only 2 years previously.

Watkins-Pitchford (Reference Watkins-Pitchford1910) developed methods and timings for killing R. appendiculatus ticks by dipping cattle in sodium arsenite solution at intervals of 3 to 5 days. This intensive regime became central to the eradication of T. parva from the southern African population of this tick. Eradication happened by a process that was probably not fully understood at the time, of killing sufficient numbers of the ticks to reduce the transmission rate from cow to cow to the level where the basic reproductive rate of T. parva was insufficient for its survival. With a residual activity for sodium arsenite of 1 day, a larva of R. appendiculatus attaching 2 days after dipping could possibly feed and detach before the next treatment at day 5, but nymphs with a minimum feeding period of 5 days could not. Such treatment could greatly diminish reproduction by the females, with their minimum feeding period of 7 days.

Modern acaricides with longer residual periods of activity on cattle would ease this dipping burden, but other problems with the cost of dipping merely once per week, the exposed immunological state of the treated cattle, and environmental contamination, would return to haunt tick controllers. An unfortunate consequence of the bureaucratic institutionalization of tick control, in most of the countries suffering from ECf, was a failure in some of them to adapt to changing circumstances and understanding. For example, into the mid-1990s it was possible to see recently built dip tanks, nominally operated under a regime to control ECf of mandatory intensive treatment of the indigenous cattle throughout every year; even stocks of arsenical acaricides were stored at some dips.

The cost and toxicity problems with acaricides stimulated research to reduce use by partial treatment of hosts at the predilection sites of ticks. Traditional hand picking of engorging females will be selective and there have long been spot treatments at favoured feeding sites, for example grease formulations on ears against R. appendiculatus. This approach can be improved by close observations of attachment behaviour. For example, A. variegatum adults crawl on the ground to approach cattle and attach at their feet before moving to their full feeding sites on the ventral surfaces of the torso. This can be exploited by using economical footbaths to apply acaricide rather than full dips (Stachurski and Lancelot, Reference Stachurski and Lancelot2006).

Opportunities to control ticks during their questing phases

Isolation as a general method of veterinary public health was adopted as pasture management for eradicating 1-host boophilid ticks. Could it be further adapted to routine control of other types of ticks? For R. appendiculatus the answer lay in studies from the eradication of T. parva using 15 months of pasture rotation against R. appendiculatus. For farmers operating under normal conditions this timing is uneconomical.

The extraordinary longevity, as adults, of these sedentary and obligate blood-feeders was investigated with experiments on their water metabolism. Multiple mechanisms to resist desiccation were revealed, as reviewed by Sonenshine (Reference Sonenshine1991). Some of these mechanisms are active, such as the absorption of water from hygroscopic salts secreted by the salivary glands onto the mouthparts (Knülle and Devine, Reference Knülle and Devine1972). Moreover, Lees (Reference Lees and Woolley1964) showed that the ability to resist desiccation declines with age. This decline is considered to be caused by depletion of energy stores of the tick (Needham and Teel, Reference Needham, Teel, Sauer and Hair1986).

Despite such adaptations, desiccation remains the principal abiotic threat to ticks. Thus for the 1-host boophilid ticks, the survival of the small larvae, as the only questing stage, appeared to be a susceptible target for control by pasture management. Larvae of these, and other genera of ticks can be seen clustering in tight clumps at the tips of grass stems, waiting for a host whilst exposed to hot sun and dry air. Harley and Wilkinson (Reference Harley and Wilkinson1971) proposed a rotation based on a group of fenced pastures, with acaricidal treatment during movement between the pastures. Managing this system proved complex due to varying survival periods of the larvae exposed to local differences in temperature and humidity (thus saturation deficit), and greatly influenced by winter and summer, wet season and dry season, aspect, and vegetation type.

Nevertheless, pasture rotation remains one of the options in an integrated control scheme (Anonymous, 1984a; Hernandez et al. Reference Hernández, Teel, Corson and Grant2000). For example, Wharton et al. (Reference Wharton, Harley, Wilkinson, Utech and Kelley1969) found that a summer rotation of 14 weeks without cattle could reduce the need for acaricidal treatments from 19 per year to 7 per year. The difficulty for farmers was to know how long the rotation should be on their pastures in a specific environment. Farmers have many other concerns about how they manage their pastures, including veterinary recommendations for control of nematodes that deposit their eggs on pasture. Elder et al. (Reference Elder, Kearnan, Waters, Dunwell, Emmerson, Knott and Morris1980) found that of the farmers who rotated their pastures for any reason, only 22% did it primarily for tick control. Farmers must maintain good nutritional quality of their pastures and farmers regard recommendations for rotations to control ticks as something that reduces the carrying capacity of their land.

Ability to predict phenology of a tick species

The adoption of methods for studying the ecology of insects (Southwood, Reference Southwood1978) to ticks has been slow, partly because studies on the transmission of pathogens by ticks have to be adjusted for their radically different feeding patterns. It is essential to know the time it takes for eggs to be laid, to develop and hatch, and then for the larvae to complete their post-eclosion development; and similarly the time it takes for an engorged larva or nymph to develop to the next instar and start questing. The rates of these developments are determined by environmental heat energy. Given suitable conditions of humidity and allowing for possible complications with diapause, observations of development times and both experimentally controlled and field temperatures, provided the information on R. appendiculatus for example (Branagan, Reference Branagan1973a, Reference Branaganb; Short et al. Reference Short, Floyd, Norval and Sutherst1989a, Reference Short, Floyd, Norval and Sutherstb). These studies were extended by Newson et al. (Reference Newson, Chiera, Young, Dolan, Cunningham and Radley1984) who constructed an age-specific life table of the survival of R. appendiculatus larvae naturally exposed on vegetation.

Another technique that can be adapted for studies of ticks is age grading. One of the major energy reserves of Ixodes ricinus is the large number of lipid vesicles within the digestive cells of the gut (Walker, Reference Walker2001). The decline of this reserve over time can be measured quantitatively by weighing before and after solvent extraction of the lipid and the results have been used to explain the seasonal succession of generations of this tick in cool temperate climates (Randolph et al. Reference Randolph, Green, Hoodless and Peacey2002). Such analysis has less relevance for populations of tropical ticks on livestock with continuously overlapping generations of all instars. However, age grading based on lipid reserves could be adapted to field studies of the longevity of infections of such ticks with pathogens, following the information on R. appendiculatus in which T. parva sporoblasts steadily declined over 20 months in their prevalence and intensity of infection (Walker and Fletcher, Reference Walker and Fletcher1985).

To predict the phenology of a tick with a distribution that ranges across 40° of latitude in the tropics requires a model that will cope with widely varied local populations. Rhipicephalus appendiculatus, as an important vector, was endowed with sufficiently widespread field studies to provide tick counts and climate data representing 11 sites for a quantitative analysis of the seasonal dynamics (Randolph, Reference Randolph1994a, Reference Randolph1997). This permitted construction of a simulation model that provides predictions for the phenology at sites with substantially different climates (Randolph and Rogers, Reference Randolph and Rogers1997). The crucial factors are: development rates dependent on temperature, behavioural diapause that adjusts seasonally the start of questing by adults, density-dependent mortality in survival stage to stage effected as the ticks feed, the effect of environmental factors on survival of questing stages, limiting conditions of microclimate to enable larvae to regain lost water, and an adjustment for successfully attaching to a host. The operation of the model is also generic; given the relevant values it will work for other tick species. Essential for the development of this type of model are detailed data on numbers of all stages of the tick during the seasons, with matching climatic data.

Seasonally adjusted timing of treatments on cattle

The phenology of livestock ticks is most distinct where one short wet season alternates with a long dry season, and away from the tropics, overlain by winter and summer effects. There are other factors controlling the phenology of ticks in addition to the basic driving of physical development of each tick through time by heat energy. The local population of a tick species is likely to be adapted to a distinct cycle of wet and dry seasons by diapause. This will match the peak period of egg laying and larval questing to the season of maximum moisture and growth of vegetation on which to shelter and quest. The tougher nymphal and adult instars will be active during the desiccating conditions later on. Timing of peak questing of feeding instars may also be adapted to the best availability of hosts.

For example the life cycle of a 3-host tick such as R. appendiculatus can be completed naturally in approximately 6 months in the wet equatorial climate of Burundi or Rwanda (Kaiser et al. Reference Kaiser, Sutherst, Bourne, Gorissen and Floyd1988; Bazarusanga et al. Reference Bazarusanga, Geysen, Vercruysse and Madder2007). Such populations will consist of continuously overlapping generations. In contrast, in Zambia, with its single rainy season and 8 months harsh dry season, the cycle of R. appendiculatus is adapted to stretch over 12 months (MacLeod et al. Reference MacLeod, Colbo, Madbouly and Mwanaumo1977). This is mediated by a behavioural diapause, induced by short day length, and entered into by adults which have completed moult and post-moult development but not become active (Berkvens et al. Reference Berkvens, Pegram and Brandt1995).

Boophilid tick species can complete their 1-host life cycle in approximately 70 days; in optimal equatorial environments they have 5 generations per year. In areas with a distinct summer – winter climate for R. microplus, there is a distinct rise in numbers of females feeding and larvae questing in the spring, with an adaptation to the timing of the wet season (Evans, Reference Evans1992).

Early research led to the proposition of adjusting the timing of treatments to be more intensive, or entirely concentrated on, the period of maximum reproductive activity. One killed engorging female R. appendiculatus represents a potential of 4000 larvae never to quest, although adverse conditions may kill many of the eggs. There is considerable scope for reducing expenditure on acaricide, and some of the large indirect costs of treatment, by using seasonal adjustments. This has been applied to control R. microplus (Sutherst et al. Reference Sutherst, Norton, Barlow, Conway, Birley and Comins1979a). The practice has been named strategic treatment (Pegram et al. Reference Pegram, Hargreaves and Berkvens1995) where applications of acaricide, at the normal interval in days for that species, start and cease at the calendar time specified using local climatic information. Clearly, good information on the seasonal variation of the target ticks in various environments is required for strategic control.

To extend this principal to advice for farmers is complicated by the counter-intuitive nature of both the tick's phenology and what is apparent to the farmer. The start of strategic treatment needs to be before any easily observable population of engorging females on the cattle. If treatment starts at the more obvious time indicated by many ticks infesting the herd, then each treatment will only reduce the population partially with sharp rebounds to high levels between treatments. The enormous value of mathematical models for overcoming problems of intuition is demonstrated here. For example, deterministic and life-table based simulation models of boophilid tick species are driven by temperature and constrained by humidity and other effects such as host resistance to feeding of ticks. These factors are entered into the computer programme optionally for different regions, climates and farming conditions (Mount et al. Reference Mount, Haile, Davey and Cooksey1991). The simulation model of Randolph and Rogers (Reference Randolph and Rogers1997) has similar potential for R. appendiculatus.

The early models and ideas on strategic treatment stimulated field validation, exemplified for R. appendiculatus by De Castro et al. (Reference De Castro, James, Minjauw, DiGiulio, Permin, Pegram, Chizyuka and Sinyangwe1997), Fivaz and De Waal (Reference Fivaz and De Waal1993) and Young et al. (Reference Young, Groocock and Kariuki1988). Strategic treatment is effective against R. appendiculatus even in equatorial areas if they have sufficiently distinct wet and dry seasons (Kaiser et al. Reference Kaiser, Sutherst, Bourne, Gorissen and Floyd1988). Unfortunately, persuading farmers to adopt strategic treatment can be difficult because they are constrained by many other factors (Cook, Reference Cook1991; Jonsson and Matschoss, Reference Jonsson and Matschoss1998).

There remain research questions for strategic treatment. If the control of ECf is the priority for strategic control in southern Africa, where there is a distinct seasonal peak of R. appendiculatus nymphs, then this instar needs to be considered. Its peak of activity in the dry season is correlated with new infections with T. parva (Mulumba et al. Reference Mulumba, Speybroeck, Billiouw, Berkvens, Geysen and Brandt2000) and this instar is an effective vector in which the low numbers of theilerial sporoblasts per nymph are probably compensated by the typical 10-fold greater number of nymphs compared to adults infesting cattle (Ochanda et al. Reference Ochanda, Young, Wells, Medley and Perry1996, Reference Ochanda, Young and Medley2003). The role of nymphs as vectors deserves greater attention, especially since they are often easier to study than adults and infections of them with pathogens can be detected by molecular methods (Bishop et al. Reference Bishop, Sohanpal, Kariuki, Young, Nene, Baylis, Allsopp, Spooner, Dolan and Morzaria1992; Ogden et al. Reference Ogden, Gwakisa, Swai, French, Fitzpatrick, Kambarage and Bryant2003).

Models of tick population dynamics are becoming more detailed biologically and thus potentially more accurate and a finer aid to understanding. Corson et al. (Reference Corson, Teel and Grant2004) in their model of R. annulatus and R. microplus, include in the list of 26 parameter values, those for 5 parameters just to describe the activity of one rate-controlling enzyme involved in the development of individual ticks. In contrast, conceptual complexity may sometimes be inappropriate in the context of control schemes where models are required urgently as pragmatic tools. Dye (Reference Dye1992), for example, has even argued for the mathematical eradication of vectors from the mechanisms of some epidemiological models designed specifically for control schemes. This is intended to make the models less demanding of data during evaluation of progress.

Prediction of where ticks are likely to occur

Knowledge of where target ticks occur is essential to plan eradication campaigns and is useful for large-scale control schemes. At the scale of the microclimatic habitats of ticks a large body of knowledge has been obtained on Ixodes and other ticks that transmit zoonotic pathogens, essential for assessing and avoiding risk of exposure. Pavlovskϊ's book (Reference Pavlovskϊ1966) on this natural nidality remains an interesting introduction to this school of thought and a more modern compilation is by Sonenshine and Mather (Reference Sonenshine and Mather1994). For this review, however, it is more effective to confine discussion of distribution to species of livestock ticks.

Ticks will only be distributed within areas where they encounter those hosts to which they are adapted for either feeding as adults or both feeding and mating. The tick's reproduction must be prolific to maintain the local populations. For A. variegatum, R. appendiculatus, R. annulatus, R. decoloratus and R. microplus these maintenance hosts are ungulates, especially members of the Bovidae. Cattle are hosts to which all the above species have become well adapted, especially the above boophilid species. Thus, to maintain the barrier zone keeping the USA free from R. annulatus and R. microplus, it is necessary to have detailed knowledge of local populations of these ticks where white tailed deer possibly act as maintenance hosts, forming a nidus or refuge for the ticks (see below).

For A. variegatum potentially spreading to the mainland Americas it was essential to know how far it might spread there in the future. This species maintains itself on cattle, sheep and goats. These domestic hosts are spread over vast and varied areas of the Americas, so the question became what else limits the distribution of this tick? An early answer came from matching the geographical characters of the known pattern of distribution of A. variegatum in Africa. These characters are the outer boundaries defined by tick collectors and within which the climate and vegetation can be described as they may affect tick survival. Computerizing this matching task using generic models was an obvious improvement on mental correlation of maps. The generic model of Sutherst and Maywald (Reference Sutherst and Maywald1985) was one of the first to do this for ticks and also insects; its prediction of the potential spread of New World screwworm in Africa contributed to the successful eradication of the infestation introduced into Libya. A further development for mapping the distribution of A. variegatum was the use of field surveys of the tick in various habitats in the Caribbean combined with analysis of satellite images of vegetation classes, as a surrogate for climate, using statistical methods such as discriminant analysis (Hugh-Jones, Reference Hugh-Jones1991).

Greater biological understanding was required. Some came from correlation of data on distributions of tick species with both host, and vegetation and climate data, to test which of these three factors is most important in determining large-scale distributions of ticks. Cumming (Reference Cumming1999, Reference Cumming2002), using datasets of more than 34 000 records in the literature of distribution of many species of ticks in Africa, concluded that, at the continental scale, climate and vegetation determine these distributions. Further, Cumming (Reference Cumming2002) using logistic regression analysis, concluded that climate, summarized as the covariance of temperature and rainfall, is the best predictor of the wide-scale distribution of ticks. However, at the time that satellites came to provide images of vegetation, suitable climate data were less easy to obtain, so the normalized vegetation difference index had become a useful surrogate for moisture at ground level. Geographical information systems for ticks, now with the possibilities of satellite-derived climate data, continue to be developed powerfully, as reviewed by Daniel et al. (Reference Daniel, Kolář, Zeman, Bowman and Nuttall2008).

The way to more accurate predictive models is a combination of detailed biological process and validation against detailed field data on climate and tick distribution. The latter continues to be a problem. Some collectors have gone about their business thoroughly, thus for R. appendiculatus in Uganda, Tanzania and Zimbabwe the many ad hoc collections have been greatly enhanced by systematic country-wide collections of specific projects, for example that of Matthyse and Colbo (Reference Matthyse and Colbo1987) in Uganda. Contiguous with some of these countries is Mozambique, for example, where this species appears almost absent (Walker et al. Reference Walker, Keirans and Horak2000) but historical scarcity of collectors there is likely to be a better explanation than unsuitable environment. Horak et al. (Reference Horak, Nyangiwe, DeMatos and Neves2009) made collections in Mozambique that now demonstrate some continuity with the well-described tick populations in South Africa. Training of local researchers to continue the work of identifying and reporting ticks (Okello-Onen et al. Reference Okello-Onen, Hassan and Essuman1999; Anonymous, 2004) needs to be promoted alongside the advancing techniques of modelling.

Influence of wild animals on tick eradication and control

It is important to know what hosts may be reservoirs of infestation during an eradication campaign, or they may spread fresh infestations or infections with pathogens into the campaign region. During the campaign against A. variegatum in the Caribbean there was exceptionally good scope for isolation of treated livestock on the small islands. Unfortunately, with these islands being close together and in frequent trading contact, there was danger of re-infestation from illicit movement of livestock and fresh cattle hides, avoiding the mandatory inspection and treatment (David Hadrill, personal communication).

The movement of birds was also a risk in the Caribbean, especially the cattle egret (Bubulcus ibis), a prodigious colonizer that had arrived from the Old World. There was considerable interest shown in these birds, which feed in cattle pastures (Alderink and McCauley, Reference Alderink and McCauley1988; Barré and Garris, Reference Barré and Garris1989; Corn et al. Reference Corn, Berger and Mertins2009). In contrast to transported livestock with engorged female ticks still attached, these egrets were found to be infested only with small numbers of immature stages. The probability of engorged larvae or nymphs, detaching from an egret at a cattle pasture, actually establishing a new infestation is a diminishing product of many small probabilities. A factor in favour of such invasion by A. variegatum is that a low level of parthenogenesis has been recorded in female ticks, without males present, fed on goats in a laboratory in Puerto Rico (Garris, Reference Garris1984). The relationship between time elapsed over long periods, repeated movements of these birds, and their levels of infestation, deserves more study.

Birds also are a working example of biological control of ticks. Domestic chickens can often be seen feeding on ticks they find on and around livestock. This is exploited by small-holder farmers as a contribution to tick control (Hassan et al. Reference Hassan, Dipeolu and Munyinyi1992). Wild oxpecker birds (Buphagus spp.) are managed in a way that increases their feeding on the ticks they find on cattle and wild animals in game reserves (Bezuidenhout and Stutterheim, Reference Bezuidenhout and Stutterheim1980).

The intractable biological problem that probably was dominant in the overall failure of the eradication campaign in the Caribbean was caused by large populations of feral donkeys and unconstrained sheep and goats on some islands, acting as maintenance hosts for A. variegatum. Culling the feral hosts was impossible and treating the ovine hosts was inefficient, so emphasis was placed on the favoured status of cattle as the maintenance host for A. variegatum. Unfortunately, some cattle were almost impossible to treat, even with pour-on acaricide, due to the meagre resources for managing the cattle (Pegram and Eddy, Reference Pegram and Eddy2002; Pegram, Reference Pegram2010).

During the later stages of the campaign against cattle ticks in the USA there were massive culls of deer to assist elimination of local foci of tick infestation (George, Reference George1989b). Now, deer in the barrier zone abutting Mexico are a continual threat to cattle northwards. This threat can be reduced by acaricide passively applied to deer through self-applicators baited with pellets of cattle-feed (Pound et al. Reference Pound, Miller, George and Lemeilleur2000).

Use of immune resistance by hosts against ticks

Exploitation of host immunity to control cattle ticks with a specific breeding programme started with the Bonsmara breed in South Africa. This trend has been best documented with the breeding of the Australian Friesian Sahiwal and similar cattle which combine selection for milk yield and drought resistance in the taurus cows and immunity to ticks in the indicus bulls (Wharton et al. Reference Wharton, Utech and Turner1970; Utech et al. Reference Utech, Wharton and Kerr1978). The latter is assessed by artificial infestations with larvae and counts of the resulting number of engorged females. These breeds are commercially successful. Up to 60% of cattle farmers in an area such as Queensland were using this host immunity within several decades of the start of the research on it (Elder et al. Reference Elder, Hass, Reid, Kearnan and Emmerson1985). Such cattle are ideal for integration in systems to control R. microplus and Babesia, which also include anti-babesial drugs and vaccines, and the synthetic antigen vaccine against R. microplus.

To exploit aggregated distribution of ticks on livestock, individual animals that are persistently heavily infested compared to the rest of the herd under the same conditions are identified by counting tick infestations. They can then either be treated more intensively, or culled from the herd more rapidly, and certainly prevented from breeding (Anonymous, 1984a). The problem with indicus breeds is that their ability to acquire resistance is coupled with lower productivity compared to taurus breeds. However, with pure indicus breeds of commercial beef cattle it can be effective to cull just the 5% of individuals with least resistance to R. microplus to achieve an increase in resistance of the herd. The equivalent figure for selection within cross-breed herds is 20% culled (Seifert, Reference Seifert1984).

Economic thresholds for treatment against ticks

Eradication and control of ticks is usually done for prevention of deaths from infection with transmitted pathogens. In addition, ticks of livestock are often pathogenic in their own right. A common example is the sight of a cluster of mated feeding pairs of A. variegatum adults on a heifer's teat, often resulting in substantial reduction of suckling. Furthermore, 3 main species of livestock ticks have been assessed in projects involving large numbers of cattle exposed to managed infestations. The direct effects on production of body mass in beef cattle, or of milk, were observed as differences in yield compared to controls with lower levels of infestation. The cause of the loss, apart from wounding and rendering susceptible to myiasis, has been shown for R. microplus to be anorexia at 65% and blood loss for 35% (Seebeck et al. Reference Seebeck, Springell and O'Kelly1971). Rhipicephalus microplus is responsible for between 0·6 g (Sutherst et al. Reference Sutherst, Maywald, Kerr and Stegeman1983) and 0·9 g (Jonsson, Reference Jonsson2006) reduction in gain of live weight for each tick that matures to a detached engorged female. Equivalent information for R. appendiculatus is 4 g per tick (Norval et al. Reference Norval, Sutherst, Kurki, Gibson and Kerr1988) and for A. variegatum is 46–61 g per tick (Pegram et al. Reference Pegram, Lemche, Chizyuka, Sutherst, Floyd, Kerr and McCosker1989). Information on reduction in milk yield has been technically difficult to obtain but with infestations of A. variegatum it can also be important. The pruritus and pain at tick bites may affect, through the generalized stress that they cause, the function of hormones of growth and lactation (Symons, Reference Symons1985).

The main purpose of this research on production losses was to define economic thresholds for treating the cattle with acaricide. Equations incorporating the full cost of treatment, the cash value of the meat or milk and a damage coefficient indicate at what level of infestation it will be profitable to treat the cattle (Sutherst et al. Reference Sutherst, Maywald, Kerr and Stegeman1983). Infestations are assessed by counts of the conspicuous female ticks engorging in their final phase of feeding. Balashov (Reference Balashov1972) demonstrated that ticks need to grow many tissues during early phases of feeding in order to acquire and accommodate the final rapid filling with blood. For a female R. microplus this final phase lasts 24 h and ticks in this condition are named standard ticks. In a region of South America where R. microplus has 4 generations per year, a typical infestation counted on one day was 53 standard ticks per cow (Brizuela et al. Reference Brizuela, Ortellado, Sanchez, Osoria and Walker1996), and in Queensland, Australia, 158 ticks (Sutherst et al. Reference Sutherst, Maywald, Kerr and Stegeman1983). Integration of economic threshold treatment with other controls is complicated by the need of farmers to focus on potentially fatal infections with transmitted pathogens. It can be combined with strategic treatments, but threshold treatment is probably under-used due to the cost of frequently counting the ticks.

The benefits and costs of tick control

Benefit to cost ratios for tick control are difficult to measure and have often been found to be poor compared to those for control of crop pests. A study on control of A. variegatum on indigenous Sanga cattle, in a traditional production system in Zambia, without ECf present but with endemic stability to babesiosis and anaplasmosis, showed that intensive acaricidal control at 36 treatments per year gave the highest productivity measured as weight gain, with a benefit:cost of approximately 7:1. In the same study, a strategic control plan of 12 treatments during the wet season only was the most beneficial, at a ratio of 20:1 (De Castro et al. Reference De Castro, James, Minjauw, DiGiulio, Permin, Pegram, Chizyuka and Sinyangwe1997).

In stark contrast, Jonsson et al. (Reference Jonsson, Davis and De Witt2001) concluded that, for control of R. microplus on dairy cattle in Queensland, the costs of acaricidal control and the losses from infestation were approximately equal. Meltzer et al. (Reference Meltzer, Norval and Donachie1995) found that acaricidal control, directed mainly at A. hebraeum on indigenous cattle in Zimbabwe, actually resulted in decreased milk production and concluded that where there is endemic stability to heartwater, babesiosis and anaplasmosis, the only form of tick control should be of the threshold type. Other studies have similarly found an inverse benefit:cost ratio for intensive control for indigenous breeds of dairy cattle in the absence of ECf (Pegram et al. Reference Pegram, James, Oosterwijk, Killorn, Lemche, Ghirotti, Tekle, Chizyuka, Mwase and Chizhuka1991; Okello-Onen et al. Reference Okello-Onen, Mukhebi, Tukahirwa, Musisi, Bode, Heinonen, Perry and Opuda-Asibo1998), but moving to positive benefit for strategic control. For indigenous cattle, reared mainly for beef, on extensive dry rangeland in East Africa and without risk from ECf, there was no reported economic benefit from weekly, 3-weekly or threshold treatment with acaricide (Tatchell et al. Reference Tatchell, Chimwani, Chirchir, Ongare, Mwangi, Rinkanya and Whittington1986).

The combination of farming type and disease threat greatly affects economic assessments. For example, where deaths from ECf are the main worry for farmers with ticky animals, then the benefit:cost of tick control can be approximately 6:1, and this will be reinforced where the profitable business is made of selling milk to increasing numbers of customers in economically developing towns and cities. Furthermore, there is additional benefit in using the current vaccine against T. parva (Mukhebi et al. Reference Mukhebi, Perry and Kruska1992, Reference Mukhebi, Kariuki, Mussukuya, Mullins, Ngumi, Thorpe and Perry1995).

Changes in land use and social factors affecting endemic stability

Comparisons with similar vector control schemes are instructive: especially against tsetse (Glossina) in Africa, because these flies have been the target of many centrally directed and wide-scale campaigns employing all options. Bourn et al. (Reference Bourn, Reid, Rogers, Snow and Wint2001) made a strong case that veterinary interventions had far less effect than anthropogenic and autonomous causes of reductions in tsetse, and thus animal trypanosomiasis. Similarly, changes to habitats made by humans improving their pastures, and obviously by arable farming, are likely to be adverse for tick survival.

The influence of land-use on how endemic stability can be managed is a topic that spans the century from the sources of the conflict between small-holder farmers and commercial cattle ranchers in southern USA, to current problems in managing babesiosis and ECf. The medical concept of endemic stability originated in studies of tick-borne pathogens, when early researchers grappled with the counter-intuitive observation of more frequent clinical cases at the margins of the tick-infested areas, or associated with partial or failing tick controls. For effective use of the concept of endemic stability, in cases where tick-borne pathogens have a carrier state, a strict distinction is essential between the varied consequences of infection with a specific pathogen, and a clinical state of disease of the livestock animal that may or may not be caused by that pathogen (Coleman et al. Reference Coleman, Perry and Woolhouse2001).

The early eradication of T. parva from southern Africa proceeded without knowledge of these subtleties, but short of eradication it is dangerous to ignore the influence of endemic stability. Attitudes changed radically after the nationally administered system in Zimbabwe of intensive and compulsory tick control, by dipping with 6000 tanks that had been built to eradicate T. parva, collapsed during the war for independence in 1965–1979 (Lawrence et al. Reference Lawrence, Foggin and Norval1980). The national herd lost its herd immunity to tick-borne pathogens, resulting in an estimated death of 1 million cattle, mainly from babesiosis and heartwater. The consequent reduction in grazing pressure allowed heavily grazed areas to become more vegetated and more suitable for survival of R. appendiculatus and A. variegatum. Control of foot and mouth disease had also been delivered substantially via the dipping system – this control was disrupted and the disease worsened. Lawrence et al. (Reference Lawrence, Foggin and Norval1980) drew the lesson that if eradication cannot be achieved at a wide area or national level then other action, especially vaccination, should be taken to maintain endemic stability.

Researchers responded to this event by closer attention to recent work on the dynamics of transmission, especially of Babesia between calves by boophilid tick species. This had been developed by Mahoney and Ross (Reference Mahoney and Ross1972) who adapted the transformative concept of Macdonald (Reference Macdonald1952) about the basic reproduction rate of Plasmodium protozoa transmitted by mosquitoes. This adaptation quantified the process of calves gaining good immunity to Babesia by exposure to infected ticks whilst protected by colostrum from their dams. Practical application of this approach was described in a field manual (Anonymous, 1984b). In this, the frequency of transmission of Babesia from larval ticks is the probability that each calf will receive infection on one day, that is: I=1 – e^−ht, where I is the proportion of calves infected, h is the daily probability of infection (inoculation rate), and t is the mean age of the calves. Smith (Reference Smith1983) expanded use of this approach to the dynamics of Babesia transmission with a computer simulation model. Under ideal circumstances this virtuous circle, of stimulation of immunity by high probability of exposure to low numbers of Babesia, could be maintained by sub-clinical carrier infections of Babesia in the cattle, transmitted through the herd by larval ticks. These larvae become infected by the transovarial route, which seems to be a difficult process for the Babesia to judge by the very low prevalence of infection in larvae that result from favourable circumstances in a laboratory. However, with 2000 larvae produced per female, there are often in a natural population of larvae sufficient of them infected to maintain endemic stability in grazing cattle.

Sserugga et al. (Reference Sserugga, Jonsson, Bock and More2003) studied how practical it is for farmers to manage the tick and Babesia populations on cattle pastures to exploit endemic stability. They assessed infections with B. bovis, B. bigemina, and Anaplasma marginale on dairy farms in Queensland and found endemic stability to be rare for all 3 parasites. The pattern of infection in cattle was poorly related to methods of tick control and the authors considered that vaccination should be the main means of inducing endemic stability to reduce disease in dairy cattle.

The question of whether host immunity confers sterile immunity against T. parva has been highly contentious from the days of eradication of T. parva through to the initial deployment of the infection and treatment vaccine. Now, the emphasis for applied research is on the role of the carrier infections which are considered typical and of central importance in endemic stability (Young et al. Reference Young, Leitch, Newson and Cunningham1986; Norval et al. Reference Norval, Lawrence, Young, Perry, Dolan and Scott1991; Latif et al. Reference Latif, Hove, Kanhai and Masaka2001).

In areas endemic for ECf, and where vaccination is available, there remain unresolved questions about what level of tick control is appropriate to reduce the costs of disease from other tick-borne pathogens and direct parasitism by ticks. Kivaria et al. (Reference Kivaria, Ruheta, Mkonyi and Malamsha2007), for example, has shown in Tanzania that treatment with acaricide can be reduced, but only slightly if good productivity is to be maintained in high yielding commercial cattle. Gitau et al. (1999) in a similar area in Kenya found clinical ECf to be uncommon on intensively farmed land. Dairy cattle of exotic breeds in these wet fertile environments are often zero-grazed, that is, kept stalled and fed the cut fodder brought to them. For such cattle in such an area, ECf is rare, and with increasing intensification of agriculture in response to human population pressure, zero-grazing will become more common. Hence the cognitive dissonance suffered by some long-experienced researchers at the sight of exotic breeds of dairy cows thriving in areas where T. parva used to threaten farmers with dead cattle. An even more emphatic finding is from Rubaire-Akiiki et al. (Reference Rubaire-Akiiki, Okello-Onen, Musunga, Kabagambe, Vaarst, Okello, Opolot, Bisagaya, Okori, Bisagati, Ongyera and Mwayi2006) who not only corroborated the above conclusions of Gitau and co-workers but, in their study in Uganda, found that zero-grazing in fertile areas reduces the risk of all species of ticks on cross-bred dairy cattle. The zero-grazing resulted in a general loss of endemic stability in the area, and this is becoming an agricultural trend. Rubaire-Akiiki et al. (Reference Rubaire-Akiiki, Okello-Onen, Musunga, Kabagambe, Vaarst, Okello, Opolot, Bisagaya, Okori, Bisagati, Ongyera and Mwayi2006) recommended that since the farmers manage these cattle without vaccination against ECf it should not be introduced because of the cost and the danger of T. parva from the live vaccine getting into the small numbers of vectors present and then causing acute disease. Hence the research imperative for synthetic antigen vaccines against these protozoan pathogens. Meanwhile the successful deployment of the live vaccine against T. parva in various farming conditions continues and has been reviewed by Di Guilio et al. (Reference Di Giulio, Lynen, Morzaria, Oura and Bishop2009).