DRm

Aa

b

c

a

ARRA

KTHHIRIR

1

s1s

0h

Veterinary Parasitology 188 (2012) 346– 354

Contents lists available at SciVerse ScienceDirect

Veterinary Parasitology

jou rn al h om epa ge: www.elsev ier .com/ locate /vetpar

ifferential feeding success of two paralysis-inducing ticks,hipicephalus warburtoni and Ixodes rubicundus on sympatric smallammal species, Elephantulus myurus and Micaelamys namaquensis

. Harrisona,∗, G.N. Robba, N.C. Bennetta, I.G. Horakb,c

Mammal Research Institute, Department of Zoology and Entomology, University of Pretoria, Hatfield, Pretoria 0028, South AfricaDepartment of Veterinary Tropical Diseases, Faculty of Veterinary Science, University of Pretoria, Onderstepoort 0110, South AfricaDepartment of Zoology and Entomology, University of the Free State, Bloemfontein 9301, South Africa

r t i c l e i n f o

rticle history:eceived 20 September 2011eceived in revised form 20 March 2012ccepted 23 March 2012

eywords:oxic paralysisost specificityost choice

xodeshipicephalus

nsectivoreodent

a b s t r a c t

Rodents are recognised as important hosts of ixodid ticks and as reservoirs of tick-bornepathogens across the world. Sympatric insectivores are usually inconspicuous and oftenoverlooked as hosts of ticks and reservoirs of disease. Elephant shrews or sengis of theorder Macroscelidea are small insectivores that often occur in sympatry with rodents insouthern Africa. Sengis are invariably parasitised by large numbers of immature ticks whilesympatric rodents are infested with very few. The reason for the difference in tick para-sitism rates between these hosts is unknown. While a number of mechanisms are possible,we hypothesised that certain tick species exhibit “true host specificity” and as such wouldonly attach and feed successfully on their preferred host or a very closely related hostspecies. To investigate this, we conducted feeding experiments using two economicallyimportant tick species, the brown paralysis tick, Rhipicephalus warburtoni and the Karooparalysis tick, Ixodes rubicundus and two sympatric small mammal species as potentialhosts, the eastern rock sengi, Elephantulus myurus and the Namaqua rock mouse, Micae-lamys namaquensis. Ticks attached and fed readily on E. myurus, but did not attach or feedsuccessfully on M. namaquensis suggesting that these ticks exhibit true host specificity. Wesuggest that a kairomonal cue originating from the odour of E. myurus may stimulate the

attachment and feeding of these ticks and that they further possess immunosuppressivemechanisms specific to E. myurus, allowing them to feed on this host species but not onM. namaquensis. This study highlights the importance of small mammalian insectivores aspotential hosts of ixodid tick species and hence their potential as reservoirs of tick-bornepathogens.

© 2012 Elsevier B.V. All rights reserved.

. Introduction

Small mammals are important hosts for the immature

tages of ixodid ticks throughout the world (Sonenshine,991). The importance of rodents in supporting these lifetages is well documented (Norval, 1979; Talleklint and

∗ Corresponding author. Tel.: +27 0713815103.E-mail address: atharrison@zoology.up.ac.za (A. Harrison).

304-4017/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.vetpar.2012.03.042

Jaenson, 1997; Clark et al., 1998) and the role of rodentsas reservoir hosts of tick-borne pathogens is similarly wellknown (Donahue et al., 1987; Randolph et al., 1999; Bownet al., 2003; Karbowiak, 2004). However, the importanceof sympatric insectivore species in supporting tick popula-tions, and as reservoirs of tick-borne pathogens, is poorly

understood. Available data suggest that insectivores maybe more important hosts for immature ticks than sym-patric rodents and that they may be important reservoirs oftick-borne pathogens. For example, in a forest in Dutchess

y Parasi

A. Harrison et al. / Veterinar

County, NY, USA, Brisson et al. (2008) concluded that 2shrew species, Blarina brevicauda and Sorex cinereus, fed35% of larval blacklegged ticks, Ixodes scapularis, and 55%of all larvae infected with the Lyme spirochaete, Borreliaburgdorferi. In contrast, white-footed mice, Peromyscus leu-copus, fed 10% of larvae and 25% of infected larvae of I.scapularis. Similar results were observed in the UK, wherecommon shrews, Sorex araneus, carried six times more I.trianguliceps larvae and twice as many I. ricinus larvae thansympatric field voles, Microtus agrestis, and had a higherprevalence of infection with Anaplasma phagocytophilum(Bown et al., 2011).

In southern Africa, rodents are typically infested withsmall numbers of ticks. In contrast, elephant shrews or sen-gis of the order Macroscelidea (small insectivores endemicto Africa) are invariably parasitised by large numbers ofimmature ticks (Fourie et al., 1992, 2005; Horak et al., 2011;Harrison et al., 2011), hosting approximately 27 specieswithin 6 genera (Fourie et al., 1995).

Harrison et al. (2011) found that the eastern rock sengi,E. myurus, had an overall mean tick burden of 399 comparedto 1 for the Namaqua rock mouse, Micaelamys namaquen-sis, in Limpopo province, South Africa. Similar differencesin tick parasitism rates have also been recorded for thesespecies at other locations in South Africa and in Zambia(MacLeod, 1970; Colbo and MacLeod, 1976; Fourie et al.,1992).

The reason for such large differences in tick parasitismrates between these hosts is currently unknown but anumber of non-mutually exclusive mechanisms may bepossible. M. namaquensis is social, living in family groups(Smithers, 1971; Choate, 1972), while E. myurus is solitary(Ribble and Perrin, 2005), therefore, M. namaquensis maybenefit from allogrooming. Differences in immunity maycontribute; some small mammal species develop immu-nity to ticks while others do not (Du Toit et al., 1994;Dizij and Kurtenbach, 1995). M. namaquensis is nocturnal(Fleming and Nicolson, 2004), while E. myurus is activeboth day and night (Ribble and Perrin, 2005). Ticks alsoexhibit daily activity patterns (Belozerov, 1982; Maddenand Madden, 2005), therefore, differences in the activitypatterns of ticks and their hosts may also affect contactrates. Moreover, ticks may exhibit “true or physiologi-cal host specificity” where ticks go through a process ofhost recognition, often mediated through specific chem-ical compounds associated with host odour, and possessmechanisms to suppress hemostasis or immune responsesof the specific host (Sonenshine, 2006).

The current study investigated differences in tick para-sitism rates between E. myurus and M. namaquensis. Giventhe pronounced differences in parasitism rates observedpreviously for these hosts, we hypothesised that tickswould exhibit “true host specificity” and would only attachand feed successfully on E. myurus but would not attach orfeed successfully on M. namaquensis. To test this hypothesiswe conducted feeding experiments using E. myurus and M.namaquensis as potential host species for 2 economically

important tick species, the brown paralysis tick, Rhipi-cephalus warburtoni and the Karoo paralysis tick, Ixodesrubicundus. Before its description by Walker and Horak,2000 (in Walker et al., 2000), R. warburtoni was variously

tology 188 (2012) 346– 354 347

referred to as an atypical strain of R. appendiculatus, R.punctatus, R. punctatus-like, R. pravus and R. pravus-like(references cited in Walker et al., 2000). Adults of thesetick species feed on sheep and goats causing toxic paral-ysis in central and southern South Africa, with thousandsof animals dying annually (Fourie et al., 1988; Spickett andHeyne, 1988). The larvae and nymphs of both species arecommonly found on E. myurus (Fourie et al., 1992, 2005).

2. Materials and methods

2.1. Initial small mammal and tick collection

Small mammals were trapped across 7 consecutivenights between 7 and 13 December 2010 at the farms“Nooitgedacht” (−29◦37′40′′S, 25◦38′20′′E) and “Langberg”(−29◦36′15′′S, 25◦28′10′′E) located in the south westernFree State province, South Africa. These locations were cho-sen as they are central to the distribution of R. warburtoniand I. rubicundus and to the distribution of toxic paraly-sis they are known to cause (Fourie et al., 1988; Spickettand Heyne, 1988; Walker et al., 2000). The farms also hadsheep and goats as livestock and contained suitable rockyhabitat for E. myurus and M. namaquensis. One hundredand twenty Sherman live traps, baited with a peanut but-ter, oats and sardine mix were set on rocky outcrops at10-m intervals. Depending on the terrain, either 2 tran-sects of 60 or 4 transects of 30 were set in parallel linesspaced 10-m apart. All traps were set after 16.00 and col-lected before 08.00 the following day. The body of eachcaptured individual was searched for the presence of tickswith particular attention given to the ears, base of the tail,and legs where ticks were found to aggregate. The remain-der of the body was searched by back combing the fur.Ticks were removed using fine forceps and placed in 70%ethanol for processing. Animals were housed in standardrodent containers, with a sawdust substrate, supplied withwater and, depending on the species, fed cat food, gratedapple and grated carrot (E. myurus) or mouse pellets (M.namaquensis). Animals were then transported to a climatecontrolled room in the Department of Zoology, Universityof Pretoria. Collections in the Free State were authorised bythe Department of Economic Development, Tourism andEnvironmental affairs under permit number 01/7675 andtheir import to Gauteng was authorised by the GautengDirectorate of Nature Conservation permit number CPB6-003501. Ticks were identified by I.G.H. using descriptionsprovided by Apanaskevich et al. (2007), Apanaskevich andHorak (2008), Theiler and Salisbury (1959) and Walkeret al. (2000), and by comparison with reference material.Once identified, they were counted and their developmen-tal stage recorded as larva or nymph. A second trappingbout took place across 5 consecutive nights between 3 and7 April 2011 at the farm “Langberg”. This was required inorder to collect live, engorged, larval ticks from E. myuruswhich could then be allowed to moult into unfed nymphsfor use in feeding experiments and to obtain additional E.

myurus to increase their sample size for experiments. Ani-mals were trapped as described above but housed in smallmammal containers with a wire mesh floor suspendedover a tray covered in moist tissue paper to facilitate the

3 y Parasi

cejsfDisitrsp(vosholaaami

2

esowaMiapRtblrsvenptmBes1aatmowE

48 A. Harrison et al. / Veterinar

ollection and survival of engorged, detaching ticks. Thedge of the tray was covered with a thin layer of petroleumelly to prevent ticks from leaving the housing area. Tis-ue paper was inspected each morning and each eveningor the presence of detached ticks and the tissue replaced.etached ticks were collected using fine forceps, placed

n glass vials plugged with moist cotton wool and kepthaded from direct sunlight. Of the 14 E. myurus capturedn the second trapping spell, 4 were transported back tohe climate room at the University of Pretoria while theemaining 10 were released at their respective trappingites. These animals were collected and transported underermit numbers 01/9031 (Free State) and CPB6-003470Gauteng), respectively. Ticks were transported to the Uni-ersity of Pretoria and housed in an acaridarium consistingf a glass tank filled to one tenth its volume with saturatedalt solution in order to maintain an approximate relativeumidity of 70% to facilitate their survival and devel-pment. The acaridarium and the small mammals wereocated in the same climate room, which was maintainedt a temperature of 23 ◦C and 12 h light (06.00–18.00),nd 12 h dark (18.00–06.00) cycle. Engorged larvae werellowed to moult to nymphs and were permitted a mini-um of 2 weeks for their mouthparts to harden before use

n feeding experiments.

.2. Tick feeding experiments

Unfed nymphs that had moulted from larvae werexamined under a dissecting microscope, identified topecies and divided into groups of 30 of either R. warburtonir I. rubicundus. In experiment 1, we hypothesised that R.arburtoni nymphs would exhibit true host-specificity and

s such would only attach and feed on E. myurus but not on. namaquensis. To investigate this, we utilised tick feed-

ng experiments and compared the number of unengorgednd engorged R. warburtoni nymphs collected from eachotential host species over the feeding period. 30 unfed. warburtoni nymphs were emptied into a petri dish con-aining water to restrict the ticks’ movement. A fine bristlerush was then used to place the ticks on the fur of the

ower back of a single E. myurus which had been manuallyestrained. The animal was then placed in a holding cageet up as described above for the collection of engorged lar-ae. This procedure was repeated for 4 males and 4 femalesach of E. myurus and M. namaquensis, thus a total of 240ymphs of R. warburtoni were used to infest each of the twootential host species. Although every measure was takeno ensure the welfare of the animals involved, 2 female E.yurus died suddenly on days 6 and 8 of experiment 1.oth animals were kept for an additional day to allow anyngorged ticks to detach at which point the bodies wereearched for ticks as described above. Similar to experiment, we hypothesised that I. rubicundus nymphs would onlyttach and feed on E. myurus but not on M. namaquensisnd used an identical protocol to experiment 1 to inves-igate this. However, this experiment was replicated for 3

ales and 2 females of E. myurus and 3 females and 2 malesf M. namaquensis, thus a total of 150 I. rubicundus nymphsere used to infest each of the two potential host species.

xperiments were run consecutively and 2 males and 1

tology 188 (2012) 346– 354

female of each host species from the first experiment werealso used in the second experiment, however there was a 2week break between experiments to ensure that any ticksnot accounted for in the first experiment did not affect thesecond experiment. To give an indication of any potentialeffects of immunity on tick-host relationships, the exper-iment was also replicated for R. warburtoni using a singlefemale each of E. myurus and M. namaquensis that had beenborn in the laboratory, and thus had not been exposed toticks previously. The lid, inside and outside of the hold-ing tank and the tissue paper was searched daily for thepresence of ticks. All ticks were collected and examinedunder the microscope to verify the species and to assessthe level of engorgement of the tick. Ticks were deemedto be “partially/fully engorged” if there was a visible pres-ence of blood in the gut, characterised by a dark mass anda swelling of the tick body. Ticks in which the presenceof blood could not be detected were designated “unen-gorged”. Experiments continued until detached engorgedticks were not recorded for 4 consecutive days, at whichpoint animals were searched manually, as described above,for any ticks that had not detached. Ticks that imbibedblood were weighed and placed in glass vials plugged withcotton wool and put in the acaridarium to allow moulting.For each engorged tick the day of detachment and timeto moult was recorded. All work was carried out underapproval from the Animal Use and Care Committee, Uni-versity of Pretoria, under permit number EC056-10.

3. Results

3.1. Field data

A total of 10 E. myurus (3 males and 7 females) and 24 M.namaquensis (11 males and 13 females) were captured andsampled for ticks during the first trapping session. Therewas a marked difference in the tick fauna and tick abun-dance between E. myurus and M. namaquensis (Table 1). E.myurus was heavily and predominantly parasitised by R.warburtoni but this tick species was almost absent from M.namaquensis. M. namaquensis was parasitised at very lowlevels by the immature stages of Haemaphysalis elliptica,Hyalomma truncatum and Rhipicephalus gertrudae, how-ever, these tick species were completely absent on E.myurus. In contrast, E. myurus was parasitised at very lowlevels by the immature stages of Amblyomma marmoreumwhile this tick was absent on M. namaquensis. During thesecond trapping session 14 animals were captured yieldingapproximately 780 newly moulted R. warburtoni nymphsand 730 newly moulted I. rubicundus nymphs as a conse-quence of the 7 days during which engorged larvae werecollected. No nymphs of other tick species were recoveredfrom larvae that had moulted.

3.2. Tick feeding experiments

3.2.1. Recovery of unengorged ticks

3.2.1.1. Experiment 1. There was a marked difference in thebehaviour of ticks in response to each host. On the firstday post infestation, 14 (5.8%) unengorged R. warburtoninymphs left the bodies of E. myurus and were recovered

A. Harrison et al. / Veterinary Parasitology 188 (2012) 346– 354 349

Table 1Mean larval and nymphal tick abundance (±SE) on Elephantulus myurus and Micaelamys namaquensis in the south western Free State, South Africa (LL = larvae,NN = nymphs).

Tick species Stage E. myurus (n = 10) M. namaquensis (n = 24)

R. warburtoni LL 195.20 ± 61.20 0.08 ± 0.06NN 18.10 ± 5.72 0

H. elliptica LL 0 0.08 ± 0.06NN 0 0.42 ± 0.17

H. truncatum LL 0 1.08 ± 0.40NN 0 0

R. gertrudae LL 0 10.96 ± 0.23NN 0 0.50 ± 0.19

A. marmoreum LL 0.20 ± 0.06 0NN 0.20 ± 0.06 0

Table 2Total numbers of engorged and unengorged Rhipicephalus warburtoni and Ixodes rubicundus nymphs recovered from Elephantulus myurus and Micaelamysnamaquensis after infestation with 30 unengorged nymphs.

E. myurus M. namaquensis

R. warburtoni, n = 240 I. rubicundus, n = 150 R. warburtoni, n = 240 I. rubicundus, n = 150

Unengorged 25 (10.4%) 9 (6.0%) 182 (75.8%) 87 (58.0%)Engorged 99 (41.3%)a 33 (22.0%) 1 (0.4%) 0 (0.0%)

rom tick).

a Excluding data from 2 deceased females = 56 (23.3%); excluding data fexcluding both groups, i.e. natural detachment from live hosts = 45 (18.8%

from the housing or under-floor tissue as opposed to 128(53.3%) recovered from M. namaquensis. This representedmean tick collections of 1.8 ± 0.6 (SE) and 16.0 ± 1.3 forE. myurus and M. namaquensis, respectively (Fig. 1). Ondays 2–5, 7 unengorged nymphs were collected from thehousing or under-floor tissue of E. myurus as opposed to47 from those of M. namaquensis, while a further 3 unen-gorged ticks were collected from E. myurus and 7 from M.namaquensis during the remaining 13 days of experiment

1. Overall, significantly more unengorged R. warburtoninymphs were recovered from M. namaquensis than from E.myurus (�2 = 80.83, df = 1, P < 0.001, Table 2). The same pat-tern was observed for R. warburtoni nymphs used to infest

Fig. 1. Numbers of unengorged Rhipicephalus warburtoni nymphs collected fromMicaelamys namaquensis on each consecutive day post infestation with 30 uneng8 n = 6, experiment 1).

s collected by manual searching at the end of experiment 1 = 88 (26.7%);

the laboratory reared, naive E. myurus and M. namaquensis.No unengorged ticks were recovered from the naive exper-imentally infested E. myurus throughout the whole studyperiod while a total of 24 (80%) unengorged R. warburtoninymphs were collected from a single naive M. namaquen-sis, 19 (60%) of which were recovered on the first day afterinfestation.

3.2.1.2. Experiment 2. A similar pattern was observed for

I. rubicundus. On the first day post infestation, 8 (5.3%)unengorged I. rubicundus nymphs were collected from thecages of E. myurus as opposed to 63 (42.0%) from thoseof M. namaquensis, representing mean tick collections of

the holding tanks and under tank substrate of Elephantulus myurus andorged ticks (M. namaquensis n = 8, E. myurus: day 1, n = 8; day 6, n = 7; day

350 A. Harrison et al. / Veterinary Parasitology 188 (2012) 346– 354

F he holdin ks, n = 5

1surfoeIt

33oewMIunrfomommatwdetTtttdt(dn

ig. 2. Numbers of unengorged Ixodes rubicundus nymphs collected from tamaquensis on consecutive days post infestation with 30 unengorged tic

.6 ± 0.4 and 11.8 ± 0.6 for E. myurus and M. namaquen-is, respectively (Fig. 2). On days 2–5 post infestation, nonengorged ticks were collected from the cages of E. myu-us while a total of 23 unengorged nymphs were collectedrom those of M. namaquensis. During the remaining 9 daysf experiment 2, a single tick was recovered from a cage ofach host species. Overall, significantly more unengorged. rubicundus nymphs were recovered from M. namaquensishan from E. myurus (�2 = 93.19, df = 1, P < 0.001, Table 2).

.2.2. Recovery of engorged or partially engorged ticks

.2.2.1. Experiment 1. Again marked differences werebserved in the engorgement of ticks in response toach host species and significantly more engorged R.arburtoni nymphs were collected from E. myurus than. namaquensis (�2 = 121.31, df = 1, P < 0.001, Table 2).

n total, 99 (41.3%) of the unengorged R. warburtonised to infest E. myurus were recovered as engorgedymphs while only a single, partially engorged nymph wasecovered from M. namaquensis. Of these 99, 23 detachedrom a deceased female on day 7, 20 from a deceased femalen day 9 while 11 were collected manually from the ani-als on day 19 of experiment 1. Thus the total number

f engorged ticks that naturally detached from live ani-als was 45 (18.8%). These ticks began to detach from E.yurus on day 7 and continued until day 14 (Fig. 3). Over-

ll the mean number of engorged R. warburtoni nymphshat detached naturally from E. myurus across days 7–14as 0.9 ± 0.6. The mean time to detachment was 10.1 ± 0.3ays with most ticks detaching on day 9. When taking allngorged ticks into account, regardless of when or howhey detached, their overall mean mass was 5.8 ± 0.4 mg.here was a significant difference in mass betweenicks that detached on different days (Kruskall–Wallisest adjusted for ties, H = 49.6, df = 7, P < 0.001) withicks that detached on the first day of detachment,ay 7, being significantly lighter (1.3 ± 0.5 mg) than

icks that detached on any other day (5.9 ± 0.7–9.2 ± 0.8means ± SE), Mann–Whitney U tests, W = 354.5–378.0,f = 7, P < 0.001–0.01). A total of 4 engorged R. warburtoniymphs were collected from the single laboratory reared,

ng tanks and under tank substrate of Elephantulus myurus and Micaelamys (experiment 2).

naive E. myurus, 1 on day 10, 2 on day 11, and 1 on day 12with a mean mass of 8.8 ± 0.6 mg. No engorged ticks werecollected from the naive M. namaquensis.

3.2.2.2. Experiment 2. In total 33 (22%) of the unengorged I.rubicundus nymphs used to infest E. myurus were collectedas engorged nymphs while no engorged nymphs wererecovered from M. namaquensis (�2 = 37.07, df = 1, P < 0.001,Table 2). Ticks started to detach from E. myurus on day 6and continued until day 10 (Fig. 4). No ticks were recoveredfrom the manual search at the termination of the experi-ment. The mean number of engorged I. rubicundus nymphsthat detached naturally from E. myurus across days 6–10was 1.3 ± 0.6. The mean time to detachment for I. rubicun-dus nymphs was 7.5 ± 0.2 days with most ticks detaching onday 8. The overall mass of ticks collected was 3.8 ± 0.2 mg.There was a significant difference in the mass of ticks thatdetached on different days (Kruskall–Wallis test adjustedfor ties, H = 13.87, df = 3, P < 0.01) with ticks on day 9 beingheavier (5.1 ± 0.1) than ticks that detached on all other days(3.1 ± 0.2–3.8 ± 0.3 (means ± SE), Mann–Whitney U tests,W = 23.5–85.5, df = 3, P < 0.01–0.05). Day 10 was omittedfrom the analysis as only one engorged nymph was recov-ered.

3.3. Moult success

Of the 99 engorged R. warburtoni nymphs collected fromE. myurus, 21 moulted successfully to the adult stage. Meantime to moult was 39.6 ± 0.4 (SE) with most ticks moultingon day 39. As the period from engorgement of nymphs of I.rubicundus until moulting varies between 2 and 9 months(Fourie and Horak, 1994) the same data were not recordedfor this tick species, however, after a period of 5 months,26 of the 33 engorged I. rubicundus nymphs collected fromE. myurus had moulted successfully.

3.4. Personal observations

A number of personal observations were made dur-ing the course of the experiments. Firstly, when infesting

A. Harrison et al. / Veterinary Parasitology 188 (2012) 346– 354 351

Fig. 3. Mean numbers of engorged Rhipicephalus warburtoni nymphs (±SE) collected from the holding tanks and under tank substrate of Elephantulusmyurus and Micaelamys namaquensis on consecutive days post infestation with 30 unengorged Rhipicephalus warburtoni nymphs (only data from ticks thatdetached naturally are presented, data from 2 deceased female E. myurus and from manual collections on day 19 are omitted. M. namaquensis n = 8, E.myurus: day 1, n = 8; day 6, n = 7; day 8, n = 6, experiment 1).

Fig. 4. Mean numbers of engorged Ixodes rubicundus nymphs (±SE) collected from the holding tanks and under tank substrate of Elephantulus myurus andengorge

Micaelamys namaquensis on consecutive days post infestation with 30 un

animals with ticks, it was observed that when the tickswere applied to the base of the fur of M. namaquensis,R. warburtoni would immediately begin to move awayfrom the site of infestation. The same phenomenon wasobserved for M. namaquensis and I. rubicundus, howeveractivity of this species was less than that of R. warburtoni.When applied to E. myurus, ticks of both species remainedstationary at the point of application and rarely moved.Secondly, the single R. warburtoni nymph that was classi-fied as “engorged or partially engorged” collected from M.namaquensis had consumed a very small amount of bloodso that a dark mass in the body was barely visible and

the body had not distended. Moreover, the exoskeleton ofthis tick had become unattached dorsally and anteriorlyin a manner not consistent with mechanical damage ormoulting.

d Ixodes rubicundus nymphs, n = 5, experiment 2).

4. Discussion

The current study highlights the importance of smallmammalian insectivores in the life cycle of ixodid ticks, agroup often overlooked with respect to their importance ashosts of immature ticks and as potential reservoirs of tick-borne pathogens. Nothing is known regarding the reservoircompetence of the Macroscelidea for tick-borne pathogensin South Africa, but many of the tick species, of which theysupport the immature stages, have been implicated in thetransmission of tick-borne pathogens (Walker et al., 2000).For example, R. evertsi evertsi, which is rarely collected

from sengis, can transmit a variety of tick-borne pathogensto livestock, such as Anaplasma marginale, Ehrlichi ovina,and Theileria equi (Walker et al., 2000). It has alsobeen implicated in the transmission of human infectious

3 y Parasi

dhtHbcrMlptsdep(p

antRhdodittototecoataco

tviFtasgtpo(uspacrt

52 A. Harrison et al. / Veterinar

iseases including the virus responsible for Crimean-Congoaemorrhagic fever and the bacterium Rickettsia conori,he cause of tick-bite fever (Walker et al., 2000). Recently,arrison et al. (2011) detected Anaplasma bovis, a tick-orne pathogen of wild and domestic stock, in a ticklosely resembling R. warburtoni collected from E. myu-us in Limpopo province, South Africa. The importance ofacroscelidea and other small mammal insectivores in the

ife cycle of ixodid ticks and as reservoirs of tick-borneathogens should be more fully investigated globally. Fur-hermore, E. myurus is a competent host of the immaturetages R. warburtoni and I. rubicundus, whose adults pro-uce a toxin inducing paralysis in sheep and goats (Fouriet al., 1988; Spickett and Heyne, 1988). It is also a com-etent host of the immature stages of Rhipicentor nuttalliFourie et al., 2005), whose adults produce a toxin causingaralysis in dogs (Norval and Colborne, 1985).

Both R. warburtoni and I. rubicundus were able to attachnd feed upon E. myurus but did not attach or feed on M.amaquensis. This is consistent with patterns observed inhe field data, where E. myurus hosted large numbers of. warburtoni larvae and nymphs while M. namaquensisosted only 2 incidental larvae. The absence of I. rubicun-us in field data collected in December was as a resultf seasonality, this tick species is not found on E. myurusuring the summer, as it undergoes diapause to avoid des-

ccation while it appears that R. warburtoni can withstandhese drier periods and is relatively abundant throughouthe year (Fourie et al., 1992, 2005). By placing ticks directlyn solitary hosts during feeding experiments, effects otherhan “true host specificity” that could affect the abundancef ticks on hosts, such as differences in habitat selection,ick and host activity patterns and allogrooming, wereliminated. Moreover, the response of both tick species toaptive reared, tick naive hosts, was almost identical to thatf hosts that had been previously exposed, suggesting thatcquired immunity did not play a part in the ability of tickso feed on E. myurus. We suggest that both R. warburtonind I. rubicundus exhibit true host specificity and that someue delivered by the host triggers attachment and feedingf these ticks on E. myurus.

A variety of host stimuli are responsible for the appe-ence response of ticks. Carbon dioxide, radiant heat,ibration, visual images, shadowing and host odour can allnduce an appetence response (Waladde and Rice, 1982).ourie et al. (1993) observed that adult female R. warbur-oni (then called R. punctatus) and I. rubicundus exhibitedppetence responses, in varying degrees, to radiant heat,hadow, carbon dioxide and host odour (in the form ofoat hair). However, nothing is known regarding the appe-ence responses of the juvenile stages of these ticks. Of allotential host stimuli, host odour is the most specific, withthers acting as general attractants mostly at long rangeWaladde and Rice, 1982). Whilst some or all of these stim-li may cause an initial appetence response it is likely thatome kairomonal aspect of host odour is responsible for thearticular cue for R. warburtoni and I. rubicundus to attach

nd to feed on E. myurus. While species-specific identifi-ation of potential mates via pheromones has been widelyesearched (Sonenshine, 2004) the use of kairomones forhe specific identification of hosts is poorly understood.

tology 188 (2012) 346– 354

Rechav et al. (1978) observed that Ixodes neitzi wouldaggregate on twigs that had been scent-marked by theirpreferred host, the klipspringer, Oreotragus oreotragus,using its pre-orbital gland and demonstrated that adultticks could detect this scent in the laboratory. Moreover,Osterkamp et al. (1999) observed that Boophilus microplus,a 1-host tick specific to cattle, was highly sensitive to thecattle-associated phenolic compounds, 1-octen-3-ol and 2-nitrophenol, in the host odour. Therefore, it is likely thatspecific chemical compounds associated with the odour ofE. myurus are detected by R. warburtoni and I. rubicundusand initiate an attachment and feeding response.

Ticks also have highly developed and diverse mecha-nisms to suppress both the innate and acquired immuneresponses of hosts on which they commonly feed(Sonenshine, 1993; Wikel, 1999). For example, the black-legged tick, Ixodes dammini can feed very successfully on itspreferred host, the white footed mouse, P. leucopus, due toits ability to evade and suppress the host immune response(Ribeiro, 1989 and references therein). In contrast, the sameticks fed on guinea pigs are rapidly rejected on repeatedexposure. It has been shown that successful feeding ofI. dammini on P. leucopus was mediated through a num-ber of pharmacologically active compounds that inhibitplatelet and neutrophil aggregation and T-cell activation. InE. myurus, the engorgement, mass at repletion and moult-ing success of immature I. rubicundus are not affected bysequential feeding challenges on the host and paralysisdoes not develop (Du Toit et al., 1994). In contrast, rab-bits on which the same numbers of ticks are fed develop aneffective immunity (Arthur M. Spickett, personal commu-nication, cited in Du Toit et al., 1994) as well as paralysis(Spickett et al., 1989). It is thus likely that R. warburtoni andI. rubicundus have evolved immunosupressive mechanismsin order to avoid rejection by E. myurus and to continuefeeding once attached. This notion is supported by the col-lection of the single R. warburtoni which had attempted tofeed on M. namaquensis. This tick imbibed a small amountof blood and then detached and died. Although the onlyexample, it may suggest that these ticks do not have theappropriate mechanisms to evade the host responses of M.namaquensis to feed successfully, even when they are ableto attach.

While the current study has provided some insightinto the mechanisms behind differences in tick parasitismrates between E. myurus and M. namaquensis, the questionremains as to why the juvenile stages of R. warburtoni andI. rubicundus are specific to E. myurus but do not, or can-not, feed on M. namaquensis. It has been noted that ticksfeeding on more primitive animals tend to be more hostspecific. For example, some species of Amblyomma feedonly on snakes, others only on lizards while others feedonly on primitive Australian monotremes and marsupi-als (Sonenshine, 1993). The Macroscelidea are believed tobe an ancient monophyletic group within the SuperorderAfrotheria (Rathbun, 2009). It may be that ticks specific tosengis have coevolved with their hosts over long periods

of time and as a result have become highly specialised asappears to be the case with ticks found on other primitiveanimals. E. myurus is an insectivore while M. namaquensisis a herbivore and as such occupy different trophic levels

y Parasi

A. Harrison et al. / Veterinar

and the chemical composition of their tissues is likely todiffer (see Kelly, 2000 for a review). For example, the bloodof E. myurus may have greater levels of nitrogen than M.namaquensis. Should this blood be more nutritious, it wouldbe advantageous for ticks to feed on those hosts at thistrophic level as opposed to those at a lower trophic levelwhich could lead to the evolution of mechanisms resultingin host-specificity.

The current study demonstrates that R. warburtoni andI. rubicundus exhibit true host specificity which is likelymediated through specific chemical compounds within theodour of the host. This has important implications for thepotential control of these tick species if these chemicalcompounds could be isolated and utilised in attractanttraps. The study also highlights the importance of smallmammalian insectivores as host of ixodid ticks and aspotential reservoirs of tick-borne pathogens of medical andveterinary importance.

Acknowledgements

This work was supported by post-doctoral research fel-lowships awarded to A. Harrison and G.N. Robb by theUniversity of Pretoria and a National Research Foundationgrant to I.G. Horak. The authors gratefully acknowledgesupport from the DST-NRF South African Research Chair ofBehavioural Ecology and Physiology awarded to N.C. Ben-nett. We kindly thank Archie and Nina du Plessis and Elsabeand Bokkie Kotze and their families for their hospitalityand access to facilities and field sites in the South West-ern Free State. We also thank Josh Sarli and Andre Prins forassistance with animal care.

References

Apanaskevich, D.A., Horak, I.G., 2008. The genus Hyalomma. VI. Systemat-ics of H. (Euhyalomma) truncatum and the closely related species, H.(E.)albiparmatum and H.(E.) nitidum (Acari: Ixodidae). Exp. Appl. Acarol.44, 115–136.

Apanaskevich, D.A., Horak, I.G., Camicas, J.L., 2007. Redescription ofHaemaphysalis (Rhipistoma) elliptica (Koch, 1844), an old taxon ofthe Haemaphysalis (Rhipistoma) leachi group from East and south-ern Africa, and of Haemaphysalis (Rhipistoma) leachi (Audouin, 1826)(Ixodida, Ixodidae). Onderstepoort J. Vet. Res. 74, 181–208.

Belozerov, V.N., 1982. Diapause and biological rhythms in ticks. In: Oben-chain, F.D., Galun, R. (Eds.), Physiology of Ticks. Pergamon Press,Oxford, UK, pp. 469–500.

Bown, K.J., Begon, M., Bennett, M., Woldehiwet, Z., Ogden, N.H., 2003.Seasonal dynamics of Anaplasma phagocytophila in a rodent-tick(Ixodes trianguliceps) system, United Kingdom. Emerg. Infect. Dis. 9,63–70.

Bown, K.J., Lambin, X., Telford, G., Heyder-Bruckner, D., Ogden, N.H., Bir-tles, R.J., 2011. The common shrew (Sorex araneus): a neglected hostof tick-borne infections? Vector Borne Zoonotic Dis. 11, 947–953.

Brisson, D., Dykhuizen, D.E., Ostfeld, R.S., 2008. Conspicuous impacts ofinconspicuous hosts on the Lyme disease epidemic. Proc. R. Soc. B:Biol. Sci. 275, 227–235.

Choate, T.S., 1972. Behavioural studies on some Rhodesian rodents. Zool.Afr. 7, 103–118.

Clark, K.L., Oliver, J.H., McKechnie, D.B., Williams, D.C., 1998. Distribution,abundance, and seasonal activities of ticks collected from rodents andvegetation in South Carolina. J. Vec. Ecol. 23, 89–105.

Colbo, M.H., MacLeod, J., 1976. Ecological studies of ixodid ticks (Acari,

Ixodidae) in Zambia. II. Ticks found on small mammals and birds. Bull.Entomol. Res. 66, 489–500.

Dizij, A., Kurtenbach, K., 1995. Clethrionomys glareolus, but not Apodemusflavicollis, acquires resistance to Ixodes ricinus L., the main Europeanvector of Borrelia burgdorferi. Parasite Immunol. 17, 177–183.

tology 188 (2012) 346– 354 353

Donahue, J.G., Piesman, J., Spielman, A., 1987. Reservoir competence ofwhite-footed mice for Lyme disease spirochetes. Am. J. Trop. Med.Hyg. 36, 92–96.

Du Toit, J.S., Fourie, L.J., Horak, I.G., 1994. Sequential feeding of Ixodes rubi-cundus on its natural host, Elephantulus myurus:effects on tick massand on engorgement and moulting success. Onderstepoort J. Vet. Res.61, 143–147.

Fleming, P.A., Nicolson, S.W., 2004. Sex differences in space use,body condition and survivorship during the breeding season inthe Namaqua rock mouse, Aethomys namaquensis. Afr. Zool. 39,123–132.

Fourie, L.J., Horak, I.G., 1994. The life cycle of Ixodes rubicundus (Acari: Ixo-didae) and its adaptation to a hot, dry environment. Exp. Appl. Acarol.18, 23–35.

Fourie, L.J., Horak, I.G., Van den Heever, J.J., 1992. The relative host statusof rock elephant shrews Elephantulus myurus and Namaqua rock miceAethomys namaquensis for economically important ticks. S. Afr. J. Zool.27, 108–114.

Fourie, L.J., Snyman, A., Kok, D.J., Horak, I.G., Van Zyl, J.M., 1993. The appe-tence behaviour of two South African paralysis-inducing ixodid ticks.Exp. Appl. Acarol. 17, 921–930.

Fourie, L.J., Du Toit, J.S., Kok, D.J., Horak, I.G., 1995. Arthropod parasites ofelephant-shrews, with particular reference to ticks. Mamm. Rev. 25,31–37.

Fourie, L.J., Horak, I.G., Marais, L., 1988. An undescribed Rhipicephalusspecies associated with field paralysis of Angora goats. J. S. Afr. Vet.Assoc. 59, 47–49.

Fourie, L.J., Horak, I.G., Woodall, P.F., 2005. Elephant shrews as hosts ofimmature ixodid ticks. Onderstepoort J. Vet. Res. 72, 293–301.

Harrison, A., Bown, K.J., Horak, I.G., 2011. Detection of Anaplasma bovisin an undescribed tick species collected from the eastern rock sengiElephantulus myurus. J. Parasitol. 97, 1012–1016.

Horak, I.G., Welman, S., Hallam, S.L., Lutermann, H., Mzilikazi, N., 2011.Ticks of four-toed elephant shrews and Southern African hedgehogs.Onderstepoort J. Vet. Res. 78, 1–3.

Karbowiak, G., 2004. Zoonotic reservoir of Babesia microti in Poland. Pol.J. Microbiol. 53, 61–65.

Kelly, J.F., 2000. Stable isotopes of carbon and nitrogen in the study ofavian and mammalian trophic ecology. Can. J. Zool. 78, 1–27.

MacLeod, J., 1970. Tick infestation patterns in the southern province ofZambia. Bull. Entomol. Res. 60, 253–274.

Madden, S.C., Madden, R.C., 2005. Seasonality in diurnal locomotory pat-terns of adult blacklegged ticks (Acari: Ixodidae). J. Med. Entomol. 42,582–588.

Norval, R.A.I., 1979. The limiting effect of host availability for the immaturestages on population growth in economically important ixodid ticks.J. Parasitol. 65, 285–287.

Norval, R.A.I., Colborne, J., 1985. The ticks of Zimbabwe, X. The generaDermacentor and Rhipicentor. Zimbabwe Vet. J. 16, 1–4.

Osterkamp, J., Wahl, U., Schmalfuss, G., Haas, W., 1999. Host-odour recog-nition in two tick species is coded in a blend of vertebrate volatiles. J.Comp. Physiol. A. Neuroethol. Sens. Neural Behav. Physiol. 185, 59–67.

Randolph, S.E., Miklisova, D., Lysy, J., Rogers, D.J., Labuda, M., 1999. Inci-dence from coincidence: patterns of tick infestations on rodentsfacilitate transmission of tick-borne encephalitis virus. Parasitology118, 177–186.

Rathbun, G.B., 2009. Why is there discordant diversity in sengi (Mam-malia: Afrotheria: Macroscelidea) taxonomy and ecology? Afr. J. Ecol.47, 1–13.

Rechav, Y., Norval, R.A.I., Tannock, J., Colborne, J., 1978. Attraction of thetick Ixodes neitzi to twigs marked by the klipspringer antelope. Nature275, 310–311.

Ribble, D.O., Perrin, M.R., 2005. Social organization of the eastern rockelephant-shrew (Elephantulus myurus): the evidence for mate guard-ing. Belg. J. Zool. 135, 167–173.

Ribeiro, J.M.C., 1989. Role of saliva in tick/host interactions. Exp. Appl.Acarol. 7, 15–20.

Smithers, R.H.N., 1971. The Mammals of Botswana. The Trustees of theNational Museums of Rhodesia, Bulawayo, Zimbabwe, 340 pp.

Sonenshine, D.E., 2006. Tick pheromones and their use in tick control.Annu. Rev. Entomol. 51, 557–580.

Sonenshine, D.E., 2004. Pheromones and other semiochemicals of ticksand their use in tick control. Parasitology 129, 405–425.

Sonenshine, D.E., 1993. Biology of Ticks, vol. 2. Oxford University Press,

New York, USA, 488 pp.

Sonenshine, D.E., 1991. Biology of Ticks, vol. 1. Oxford University Press,New York, USA, 472 pp.

Spickett, A.M., Heyne, H., 1988. A survey of Karoo tick paralysis in SouthAfrica. Onderstepoort J. Vet. Res. 55, 89–92.

3 y Parasi

S

T

T

Ixodidae): A Guide to the Brown Ticks of the World. Cambridge Uni-

54 A. Harrison et al. / Veterinar

pickett, A.M., Elliot, E.G.R., Heyne, H., Neser, J.A., 1989. Paralysis of labora-tory rabbits by nymphae of Ixodes rubicundus, Neuman 1904 (Acarina:Ixodidae) and some effects on the life cycle following feeding underdifferent temperature conditions. Onderstepoort J. Vet. Res. 56, 59–62.

alleklint, L., Jaenson, T.G.T., 1997. Infestation of mammals by Ixodes rici-

nus ticks (Acari: Ixodidae) in south-central Sweden. Exp. Appl. Acarol.21, 755–771.

heiler, G., Salisbury, L.E., 1959. Ticks in the South African zoologi-cal survey collection—Part IX—The Amblyomma marmoreum group.Onderstepoort J. Vet. Res. 28, 47–117.

tology 188 (2012) 346– 354

Waladde, S.M., Rice, M.J., 1982. The sensory basis of tick feeding behaviour.In: Obenchain, F.D., Galun, R. (Eds.), Physiology of Ticks. PergamonPress, Oxford, UK, pp. 71–118.

Walker, J.B., Keirans, J.E., Horak, I., 2000. The Genus Rhipicephalus (Acardi,

versity Press, Cambridge, UK, 643 pp.Wikel, S.K., 1999. Tick modulation of host immunity: an impor-

tant factor in pathogen transmission. Int. J. Parasitol. 29,851–859.