Tick – B. thuringiensis interactions

Since ticks ingest primarily host blood, it seems unlikely that B. thuringiensis, which attacks the midgut of insects, would successfully enter and cause mortality in ticks. However, Hassanain et al. (1997), performing Petri dish tests, found that three commercial varieties of B. thuringiensis (B. t. kurstaki, B. t. israelensis and B. t. thuringiensis) produced mortality when sprayed on unfed or engorged adults of Argas persicus or Hyalomma dromedarii. B. t. kurstaki was the most pathogenic (LC₅₀ 215–439 μg/ml on day 5 PI for Argas and 1200–2344 μg/ml for Hyalomma). The susceptibility of the eggs of both tick species increased with increasing time after oviposition. Zhioua et al. (1999a,b) exposed Ixodes scapularis to B. t. kurstaki by dipping engorged larvae for 30 seconds in spore suspensions, and reported pathogenicity with an LC₅₀ of 1×10⁷ spores/ml.

The crystalline δ-endotoxin of B. thuringiensis is produced during sporulation and disrupts insect midgut walls (Gill, Cowles & Pietrantonio, 1992). Unlike spores, the δ-endotoxin shows specificity to particular insect taxa (Flexner & Belnavis, 2000), therefore it is unlikely that the δ-endotoxin kills ticks by causing toxaemia. Furthermore, the δ-endotoxin of B. thuringiensis subsp. kurstaki has to be activated by an alkaline pH and specific proteases. The extent to which the toxin will be activated in ticks with a midgut pH of 6·8 and intracellular cathepsins (pH optimum 3) (Sonenshine, 1991; Ramamoorthy & Scholl-Meeker, 2001) requires additional investigation.

Ticks ingest host body fluids through their mouthparts, but they also ingest water vapour, through both mouthparts and cuticle for hydration (Knulle & Rudolph, 1982). One mechanism involves a highly hygroscopic fluid secreted by the tick that is hydrated by water vapour from the surrounding atmosphere, and is then re-ingested into the tick. Therefore, it is plausible that spraying or dipping ticks in highly concentrated bacterial solutions could result in tick mortality if sufficient fluid was ingested to elicit pathogenic effects on the midgut. However, other mechanisms of pathogenicity are also plausible, including: the actions of other toxins such as B. thuringiensis exotoxins (Sebesta et al. 1981); negative effects associated with bacterial invasion of the haemocoel (Dubois & Dean, 1995); and the physical blocking of spiracles or other openings by bacterial spores.

B. thuringiensis products have proved to be the most popular biocontrol agents in agriculture (Navon, 2000). However, because of the specific biology of ticks and the little information available, further research is required to elucidate the mechanism of bacterial pathogenicity to ticks. The mode of pathogenicity influences potential efficacy, as well as the design of delivery systems of bacterial pathogens for tick control.

Prevalence of tick – fungus interaction

In nature, 20 species of fungi have been reported to be associated with ticks. Some 13 species of ticks from 7 genera were found to be infected by fungi (Kolomyetz, 1950; Samsinakova, 1957; Steinhaus & Marsh, 1962; Cherepanova, 1964; Krylov, 1972; Samsinakova et al. 1974; Estrada-Pena, Gonzales & Casasolas, 1990; Mwangi, Kaaya & Essumen, 1995; Zhioua et al. 1999a; Guerra et al. 2001). Ticks collected in North East USA were infected primarily with Verticillium spp. and Beauveria bassiana; 10 species were isolated in Europe and 3 in Africa (Cherepanova, 1964; Samsinakova et al. 1974; Mwangi et al. 1995; Kalsbeek, Frandsen & Steenberg, 1995; Zhioua et al. 1999a). Of the engorged female B. microplus ticks collected from soil in Brazil, 24·5% were contaminated with B. bassiana, 10% with M. anisopliae (Da Costa et al. 2001), and 22% of the R. sanguineus nymphs were contaminated with fungi from 5 genera (Guerra et al. 2001). The percentage of ticks infected by fungi in collections from natural areas varies considerably, largely according to the stage and species of tick and to the ecological conditions at the sample sites. For instance, 7·5% of the adult Ixodes ricinus collected in central Europe during winter time were infected by fungi, compared to over 50% of ticks collected during the summer (Samsinakova et al. 1974). In northern Europe, between 6 and 32% of engorged I. ricinus females were infected by fungi in nature, depending on the season (Kalsbeek et al. 1995). Only 1·7% of the engorged Rhipicephalus appendiculatus females collected in Kenya died from fungal infection (Mwangi et al. 1995), and in the northeastern USA 4·3% of the collected unfed female I. scaplularis were infected (Zhioua et al. 1999a). In nature, a higher percentage of adult ticks seems to be infected by fungi than their pre-imaginal stages and engorged females seem to be most readily infected (Kalsbeek et al. 1995; Zhioua et al. 1999a).

Laboratory assays

When various fungal genera, species and strains were tested under optimal laboratory conditions, M. anisopliae and B. bassiana exhibited the strongest anti-tick pathogenicity (Guangfu, 1984; Gindin et al. 2001; Samish et al. 2001). In most cases, M. anisopliae strains were more virulent than those of B. bassiana (Mwangi et al. 1995; Castineiras et al. 1987; Barci, 1997; Kaaya & Hassan, 2000; Sewify & Habib, 2001; Zangi, 2003). Under laboratory conditions, at least 15 ixodid tick species and two argasid ticks were found susceptible to fungi (Samish et al. unpublished observation). Studies comparing the susceptibility of unfed larvae of Boophilus annulatus, Hyalomma excavatum and R. sanguineus to 12 fungal strains (from five species of fungi) gave the general impression that Boophilus and Hyalomma larvae showed similar susceptibility while Rhipicephalus larvae were more resistant. In contrast, engorged H. excavatum females were far more resistant to entomopathogenic fungi than females of the other two tick species (Gindin et al. 2003). Tick eggs, in contrast to many insect eggs, are highly susceptible to fungi and up to 100% of the eggs exposed to fungi under laboratory conditions did not hatch (Boichev & Rizvanov, 1960; Gorshkova, 1966; Castineiras et al. 1987; Monteiro et al. 1998a,b; Bittencourt, Massard & Lima, 1994a; Kaaya, 2000; Paiao, Monteiro & Kronka, 2001; Gindin et al. 2003). Fungal infection of engorged female ticks often resulted in longer periods of pre-oviposition, oviposition time, egg-incubation, and egg-hatching of the egg mass, as well as in lowered egg production (Gorshkova, 1966; Bittencourt, Massard & Lima, 1994b; Barci, 1997; Gindin et al. 2001). This suggests that there is a relatively long-lasting sub-lethal influence of the fungi on their tick hosts.

Comparison between the susceptibility of the unfed stages of R. appendiculatus and Amblyomma variegatum or of H. excavatum and R. sanguineus demonstrated decreasing susceptibility to fungi in progression through the larval, nymph and adult stages (Kaaya, 2000; Samish et al. 2001; Gindin et al. 2003), but unfed I. scapularis larvae were less susceptible than unfed adults (Zhioua et al. 1997). Unfed larvae and nymphs seem to become more resistant to M. anisopliae after engorgement (Reis et al. 2001; Gindin et al. 2003).

Fungi take several days to kill ticks. For instance, the LT₅₀ of M. anisopliae (at 1×10⁷ spores/ml) for eggs, unfed larvae and engorged females of B. annulatus, H. excavatum and R. sanguineus generally ranged from <3 to 6.5 days; in some cases it extended to weeks. Engorged larvae of H. excavatum and R. sanguineus took twice as long to die than unfed larvae of the same species, and it took over 3 weeks for 50% of unfed adults to be killed by these fungi (Gindin et al. unpublished observation).

Field trials – off host

A 0·3 ha Abies procera plantation in Denmark was sprayed with M. anisopliae (BIPESCO strain) spores (4–10×10¹⁰ spores/m²). Of the 67 unfed Ixodes ricinus ticks collected 2 weeks after treatment, 57% were contaminated with M. anisopliae (Nielsen et al. personal communication). Initial field trials used a commercial anti-termite mycoside based on M. anisopliae (ESC-1 strain, Bio-Blast Biological Termicide) against questing I. scapularis adults (spraying 4–6×10¹⁰ spores/m²). Within 4 weeks, 53% of the adults were killed (Benjamin, Zhioua & Ostfeld, 2002). When potted grass with adult or nymphal R. appendiculatus was sprayed with spores (1×10⁹ spores/ml in water) and kept in the field, the mortality of ticks sprayed with B. bassiana was 96 and 36% for nymphs and adults, respectively; if sprayed with M. anisopliae the mortality reached 76 and 64% for nymphs and adults, respectively. The two fungal species together killed over 99% of the larvae. Similar results were obtained with Amblyomma variegatum (Kaaya & Mwangi, 1995; Kaaya, Mwangi & Ouna, 1996).

In Kenya, 5 acre paddocks were seeded with R. appendiculatus larvae and sprayed once a month with M. anisopliae or B. bassiana spores in an aqueous solution (1·2×10⁹ spores/m²). During the rainy season the fungi had no effect upon the abundance of the ticks on cattle kept in the paddocks, but during the 3 months after the end of the rainy season B. bassiana and M. anisopliae reduced the tick population by 80 and 92%, respectively, compared with the control (Kaaya, 2000; Kaaya, Samish & Glazer, 2000).

Sack-cloth covering poultry houses that were heavily infested with the soft tick Argas persicus was sprayed with an M. anisopliae spore suspension (5×10⁷ spores/ml in water with Tween 80 and 4% oil) and covered with a plastic sheet to maintain high humidity. The tick population was reduced by 53% within a week and after 3 weeks no live ticks could be found (Sewify & Habib, 2001). When two commercial B. bassiana mycoinsecticides (registered for use against pests of ornamentals and turf) were tested against I. scapularis nymphs in a residential area, the tick population was reduced by 59–89%, and some 80–90% of the recovered nymphs developed mycoses (Stafford & Kitron, 2002). Ticks inoculated with B. bassiana were released in a natural D. variabilis microhabitat from which live ticks were collected one year later. Fungal growth was observed after incubation of the ticks under laboratory conditions (Lucas, Fielden & Hererra, unpublished observations). Spraying of a paddock containing B. microplus larvae with M. anisopliae spores resulted in fewer larvae one week after treatment compared to the untreated paddock (Bittencourt, 2000).

Field trials – on host

Spraying M. anisopliae (1×10⁸ spores/ml in water with Triton X-100) on gerbils one day after they had been infested with R. sanguineus nymphs reduced the number of engorged nymphs dropping off by 73% (unpublished observation). Spraying M. anisopliae spores (1×10⁸ to 1×10⁹ spores/ml in water) on cattle infested with B. microplus or B. decoloratus ticks at various development stages generally caused insignificant or low reductions (up to 50%) of the on-host tick population. However, up to 79% of the females collected from the sprayed cattle died in the laboratory, and their egg mass weight was reduced by up to nearly 50% (Castro et al. 1997; Correia et al. 1998; Bittencourt et al. 1999; Kaaya & Hassan, 2000). When B. microplus-infested cattle was treated one or four times with Verticillium lecanii (3·5×10⁷/ml, 5 l per head), the number of ticks was reduced by 48–79% or 94–99%, respectively (Camacho et al. 1998). In other trials, adult R. appendiculatus ticks were allowed to feed in ear bags on rabbits or cattle and were sprayed with M. anisopliae spores at dosages of 2·5×10⁸ or 1×10¹⁰ spores per ear, respectively; 30% of the ticks on rabbits and 83% of those on cattle died. In addition, of the eggs laid by the Beauveria-treated females, none of those on the rabbits and only 48% of those on cattle hatched (Kaaya et al. 1996). Application of a gel formulation containing B. bassiana spores to the ears of horses reduced Anocentor nitens tick populations by 50% (Bittencourt et al. unpublished observation). When B. bassiana and M. anisopliae spores in water with Triton X-100 were sprayed on thoroughly washed ears of cattle, live B. bassiana spores were recovered up to 1 week and those of M. anisopliae up to 3 weeks after application (Kaaya et al. 1996). However, fungal spores suspended in water and sprayed on cattle were found to be virulent to Boophilus ticks for less than a week (Castro et al. 1997; Bittencourt et al. 1999), and most deaths of feeding R. appendiculatus occurred 5–10 days after fungal infection (Kaaya et al. 1996).

In several experiments live ticks were collected from a fungus-infested field or from their vertebrate hosts and incubated under optimal laboratory conditions. A high percentage of these ticks died and fungi grew on many of them even though low tick mortality was observed in the field or on hosts in the same experiments. This suggests that under natural conditions the conidia often adhere to the tick integument, at least for a short time, but in many cases they do not germinate or penetrate the cuticle, and/or cause only a sub-lethal infection under field conditions.

Formulation

In most laboratory tests the spores were suspended only in water with a small amount of dispersing agent. Kaaya & Hassan (2000) compared the anti-tick effect of a suspension of M. anisopliae spores in water with 1% Tween-80 with or without 15% peanut oil. Unfed nymphal and adult ticks were immersed in the suspension and then kept outdoors on vegetation. The mortality of A. variegatum treated with the oil suspension was 30% higher for nymphs and 2·7 times higher for adults, and for R. appendiculatus 15% higher for nymphs and 4·3 times higher for adults than in applications with the water carrier without oil. The influence of the oil itself on the ticks has still to be clarified. In another trial, a powder formulation was prepared by mixing 1 volume of M. anisopliae spores with 9 volumes of various powders and was tested against R. appendiculatus adults feeding on cattle in ear-bags. The spore mixture with millet powder caused 100% mortality, maize 79%, sorghum 64% and starch 53% (Kaaya & Hassan, 2000). Treating A. nitens-infested ears of horses with a B. bassiana suspension in water resulted in less than 20% tick mortality, while their application in a polymerized pulp gel resulted in more than 50% mortality (Bittencourt et al. unpublished observation).

Clearly, the formulation in which the spores are applied is crucial to the level of control obtained with fungus-based anti-tick compounds, but very little has been published as yet on the subject. Much information in formulating fungi for use against free-living ticks can be obtained from experiments on fungal spores for the control of insect pests, but considerable research is still required to develop satisfactory formulations for the control of ticks while they are feeding on vertebrate hosts. Finally, additional research is needed to find more virulent and specific fungal strains.

Tick–nematode interactions

EPNs penetrate engorged female B. annulatus ticks almost solely via the anus or genital pore (unpublished observations). Heterorhabditid nematodes killed engorged B. annulatus females in Petri dishes after less than 2.5 h of exposure, whereas steinernematid nematodes needed more than 4 h to penetrate into ticks (Glazer Alekseev & Samish, 2001). The injection of a single heterorhabditid nematode into a tick can cause mortality (Glazer et al. 2001). The dosages of EPNs needed to kill 50 or 90% of ticks are comparable to that used commercially in the control of insect pests of plants, but the time required to kill ticks is often relatively long (Table 1).

Tick mortality caused by EPNs seems to be due to the rapid proliferation of the nematode symbiotic bacteria within the ticks, since the nematodes do not go through their natural cycle within ticks, and most infective juveniles die shortly after entry (Mauleon, Barre & Panoma, 1993; Samish, Alekseev & Glazer, 1995; Hill, 1998; Kocan et al. 1998b; Hassanain et al. 1999). In vitro experiments demonstrated that tick hemolymph hinders the growth of EPNs (Zangi, 2003), but the reason(s) for nematode mortality within ticks is/are not yet fully understood. Interestingly, when the cuticle of I. scapularis was physically slit before nematode infection, the nematodes S. carpocapsae and S. glaseri reproduced successfully (Zhioua et al. 1995).

Laboratory assays

Ticks exposed to nematodes in Petri dishes lined with moist filter paper showed that six genera of Ixodidae and two genera of Argasidae were readily killed by entomopathogenic nematodes (Table 1). Only B. microplus appeared to be resistant to nematode attack (Mauleon et al. 1993). Various tick stages differ substantially in their susceptibility to EPNs, with the fully engorged female ticks being most susceptible and the pre-imaginal stages least sensitive. During feeding, ticks were highly resistant to EPNs and their eggs were totally resistant. Unfed female ticks were killed up to six times quicker (LT₅₀ 1 d for R. bursa) than engorged ticks (6 d for R. bursa) (Samish et al. 2000a,b; Table 1). This may be connected to the strong anti-bacterial activity of the tick haemolymph (Ceraul et al. 2002).

The many strains of more than 25 nematode species from the families Heterorhabditdae and Steinernematidae display a wide range of characteristics, enabling them to adjust readily to the various different pests in different ecological niches. The 42 nematode strains tested for anti-tick activity showed varying degrees of virulence (Table 1). In laboratory tests, heterorhabditid nematodes were generally more virulent to ticks than steinernematids (Mauleon et al. 1993; El-Sadawy & Habeeb, 1998; Hill, 1998; Hassanain et al. 1999; Glazer et al. 2001). Nematode strains virulent to one tick stage of one species were found, in most cases, to be also highly virulent to other tick species and stages (El-Sadawy & Habeeb, 1998; Hassanian et al. 1999; Samish et al. 1999a,b).

Field trials

The virulence of nematode strains to ticks as measured in Petri dish tests in the laboratory often differs considerably from that measured in field trials when the nematodes are sprayed on soil. The difference could result from the specific behaviour of each nematode strain in the soil environment, e.g. how deep in the soil they prefer to live, how they react to UV irradiation, low moisture or other factors.

Among nine strains of three nematode species tested against B. annulatus ticks on soil, under simulated field conditions, the commercial Steinernema carpocapsae Mexican strain was the most effective: a nematode dosage of 50/cm² of this strain killed 100% of the engorged females with a LT₅₀ of less than 5 days (Samish et al. 1999a).

Environmental conditions strongly influence the pathogenicity of nematodes to ticks. Soils with a high silt concentration or with more than 25% (v/v) manure were found to reduce the anti-tick activity of EPNs relative to that measured on clean sandy soil (Samish et al. 1998, 1999a). The anti-tick activity of EPN strains also varied with temperature (Zhioua et al. 1995): the optimal activity for most strains was 25–28 °C, although some displayed far wider ranges of activity (e.g. 18–34 °C; unpublished observations). EPNs sprayed on humid sandy soil 3 days before ticks were added killed 100% of the ticks, whereas EPNs sprayed on soil containing 25% (v/v) cattle manure resulted in 45% mortality, and on soil containing 40–50% silt only 25% mortality (unpublished observations). EPNs sprayed on soil covered with leaf litter or grass were highly efficient in killing ticks compared with those sprayed on uncovered soil. Shading the soil with nets (10–90% shade) and irrigating it twice a day increased tick mortality from nematode attack in comparison with that on unshaded soil or on plots irrigated only once every 2 days (Zangi, 2003).

Nematodes are potentially useful tools for tick control because: (1) engorged ticks are susceptible to some EPNs and also reside in locations that are preferred by many nematode strains; (2) immobile ticks attract mobile nematodes; and (3) spraying or irrigation can be easily used to apply nematodes to a tick habitat. However, the use of nematodes may be limited to defined ecological niches because their pathogenicity is reduced by low humidity or temperature, high concentrations of manure or silt, and by differences in the susceptibility among the various tick stages and species. The wide genetic variation found among the many nematode strains, and presumably in strains yet to be found, means that genetic manipulation of nematodes could increase the range of ecological conditions in which they could be successfully applied against ticks. The development of improved formulations is also important. Finally, in-depth studies are needed to elucidate the interactions between nematodes and ticks under field conditions.

Chickens

Chickens (Gallus gallus) confined with cattle in Africa were reported to ingest an average of 338 ticks per bird during 5·5 h. Other experiments found that the birds ate from 9·7 to 81 ticks per bird per hour of foraging. At high tick concentrations, an average of 69% of the ticks were consumed by chickens (Hassan et al. 1991; Hassan, Dipeolu & Munyinyi, 1992; Dreyer, Fourie & Kok, 1997). Chickens are neither tick-specific predators nor obligatory predators, therefore their consumption of ticks depends largely on alternative food availability and the density of the tick population. Thus, chickens are unlikely to reduce tick densities below a certain level. Nevertheless, chickens maintained in any case on small mixed farms can help to reduce tick populations at nearly no cost.

Oxpeckers

Buphagus africanus (the yellow-billed oxpecker, YBO) and B. erythrorhynchus (the red-billed oxpecker, RBO), both native to Africa, are the only birds known to feed specifically on ectoparasites, especially ticks. Oxpecker populations have decreased along with reductions in the numbers of game animals, increased use of bird-poisoning acaricides and possibly also decreased tick populations (Robertson & Jarvis, 2000; Van Someren, 1951; Stutterheim & Stutterheim, 1980; Stutterheim & Brooke, 1981). However, oxpecker reintroduction efforts, an increase in game animals and the introduction of safer acaricides appear to have resulted in increases in these bird populations during the last two decades.

Stomach contents of captured oxpeckers included 16 to 408 ticks per bird (Van Someren, 1951; Bezuidenhout & Stutterheim, 1980). The consumption by young RBOs of larval, nymphal and adult Boophilus ticks averaged 1176, 1549 and 1293, respectively, during 6–7 days of exposure to tick-infested cattle. This result was extrapolated to adult RBOs as the equivalent of 12500 larval or 98 engorged B. decoloratus females. YBOs can consume about 10% more ticks than RBOs (Bezuidenhout & Stutterheim, 1980; Stutterheim, Bezuidenhout & Elliott, 1988).

Oxpeckers are visual predators, first plucking the engorged females, then searching large body areas and scissoring and eating the smaller tick stages. YBOs prefer foraging on buffaloes and white rhinos, whereas RBOs prefer other ungulate hosts. Oxpeckers prefer feeding on weak mammals and will feed repeatedly on specific individuals within the same herd, with preference for the hosts with most ticks (Van Someren, 1951; Mooring & Mundy, 1996). Buphagus birds may also feed on skin, pieces of meat and blood of the mammalian host; thus they may enlarge wounds or even open partially healed wounds, although most publications suggest that such behaviour is quite rare (Bezuidenhout & Stutterheim, 1980; Stutterheim et al. 1988; Weeks, 1999). The YBO was found to be more aggressive than the RBO.

Several programmes have attempted to reintroduce oxpeckers artificially into areas from which they had previously disappeared (Davison, 1963; Couto, 1994; Grobler, 1976, 1979). Oxpeckers are difficult to propagate in captivity and reintroduction programmes were based on translocation of birds that were captured in the wild. When enough birds were translocated, the oxpeckers established themselves successfully, and calf mortality from tick burdens was reduced (Couto, 1994).

For a successful oxpecker introduction programme, a number of problems must be considered (Mundy & Cook, 1975; Couto, 1994; C. Foggin personal communication). The tick population should be monitored to assess the importance of the birds and the success of the introduction programme. A suitable location to capture the oxpeckers is needed (e.g. where they feed on tame calves) – mist nets can be used to capture the birds. Oxpeckers are social birds, therefore a minimum of 20 birds is recommended for translocation. It is important to overcome panic among cattle and game animals in the new location when they first encounter the oxpeckers. The new release area should have a suitable climate and should be at least 10 km away from areas treated with bird-poisoning anti-ectoparasitic compounds. Amitraz or pyrethroid-based dips do not seem to be harmful. The new area should include at least 500 head of game animals or of undipped domestic animals, grazing on at least 3000 ha.

Oxpeckers seem well suited to play an important role in IPM programmes for the control of ticks. However, their major value will be as only one part of a broader integrated programme, because substantial tick populations are known to remain in areas populated with oxpeckers (Masson & Norval, 1980; Norval & Lightfoot, 1982).