Seasonal infestation of donkeys by lice: Phenology, risk factors and management

V

Sf

La

b

a

ARRA

KPEEPDB

1

lipwa

SB

0

ARTICLE IN PRESSG ModelETPAR-7219; No. of Pages 7

Veterinary Parasitology xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Veterinary Parasitology

jo u r nal homep age: www.elsev ier .com/ locate /vetpar

easonal infestation of donkeys by lice: Phenology, riskactors and management

. Ellsea,∗, F.A. Burdenb, R. Walla

Veterinary Parasitology and Ecology Group, School of Biological Sciences, University of Bristol, UKThe Donkey Sanctuary, Sidmouth, Devon, UK

r t i c l e i n f o

rticle history:eceived 7 March 2014eceived in revised form 7 April 2014ccepted 11 April 2014

eywords:hthirapteraquinectoparasitesediculosisonkeysovicola

a b s t r a c t

A longitudinal study was undertaken over a 21 months period to examine the sea-sonal abundance of lice infesting donkeys, the risk factors which predispose donkeys toinfestation and the effectiveness of louse management. All the lice seen were Bovicola (Wer-neckiella) ocellatus. A strong seasonal pattern, which was correlated with mean monthlytemperature, was observed with higher prevalence and intensity in the cooler, wintermonths (October–March). Overall infestation in these animals was over-dispersed, suggest-ing that some individuals are strongly predisposed to infestation. Donkey age and mean hairlength were characteristics which affected louse prevalence: older and younger donkeysand donkeys with longer hair harboured the highest numbers of lice. However, the practiceof coat-clipping, to reduce the infestation, resulted in a lower louse prevalence only in thesummer, suggesting that clipping is not an effective form of louse control in cooler months.Higher louse burdens were associated with larger areas of visible excoriation and hair dam-age, suggesting that B. ocellatus does adversely impact animal welfare. However, the abilityof animal carers to estimate louse presence or absence accurately on an individual donkey
was not sufficiently high to allow targeted selective treatment of heavily infested animalsto be employed effectively. As animals are housed in closed herds these findings suggestthat clipping in the summer and treating all animals with insecticide in late autumn, priorto turn-in may be an effective louse management strategy.
© 2014 Elsevier B.V. All rights reserved.

. Introduction

Infestation by lice, known as pediculosis, can affect allarge mammals and birds of domestic and agriculturalmportance. Heavy burdens of lice are associated with

Please cite this article in press as: Ellse, L., et al., Seasonal infmanagement. Vet. Parasitol. (2014), http://dx.doi.org/10.1016/j

ruritus and excoriation which result in dermal lesionshich can significantly reduce productivity, devalue hides

nd fleeces as well as posing a significant threat to

∗ Corresponding author at: Veterinary Parasitology & Ecology Group,chool of Biological Sciences, University of Bristol, Woodland Road, BristolS8 1UG, UK. Tel.: +44 1179287489; fax: +44 1173317985.

E-mail address: [email protected] (L. Ellse).

http://dx.doi.org/10.1016/j.vetpar.2014.04.012304-4017/© 2014 Elsevier B.V. All rights reserved.

animal welfare (Heath et al., 1995; Wilkinson et al.,1982). These negative impacts are associated with bothhaematophagous sucking lice and scurf-feeding chewinglice.

Several factors influence the burden of lice, with aseasonal pattern of louse abundance often reported. In tem-perate climates, the burden of the sheep chewing louse,Bovicola in sheep and cattle is highest in the winter months(Craufurd-Benson, 1941; James, 1998; Murray and Gordon,1969). Seasonal sampling of feral pony hides from equine

estation of donkeys by lice: Phenology, risk factors and.vetpar.2014.04.012

abattoirs in south west England revealed a similar trend inBovicola equi (Denny), with the highest prevalence of 60%of the hides sampled being infested in February (Gawleret al., 2005). However, in sheep kept in Perth, Australia this

dx.doi.org/10.1016/j.vetpar.2014.04.012


http://www.sciencedirect.com/science/journal/03044017

http://www.elsevier.com/locate/vetpar

mailto:[email protected]


IN PRESSG Model

Parasitology xxx (2014) xxx–xxx

Table 1Louse abundance scores used to quantify infestation by Bovicola (Wer-neckiella) ocellatus on donkeys.

Abundance score Observation

0 No lice seen in any of the five partings

At each infestation site the hair length was recorded tothe nearest 2 mm (±1 mm). The age of each individual ani-mal was obtained from a database which contained clinicaland husbandry details for every animal. Where the age was

Table 2Excoriation coverage scores, defined by the area of a donkey that wasshowing signs of hair derangement or shortening of the hair fibre due to

ARTICLEVETPAR-7219; No. of Pages 7

2 L. Ellse et al. / Veterinary

pattern was not observed, the burden of lice continued toincrease throughout the austral summer months and onlydecreased after shearing (Wilkinson et al., 1982). Louse sus-ceptibility and patterns of changing louse intensity havebeen attributed to climatic factors such as ambient temper-ature (Craufurd-Benson, 1941; James, 1998; Murray, 1968),rainfall (Murray, 1968) and solar radiation (Matthysse,1946) as well as individual animal characteristics, such asbreed and hair length (James, 1998; Wilkinson et al., 1982).Animal husbandry factors such as shearing/clipping (Allenand Dicke, 1952; Murray, 1968; Wilkinson et al., 1982) andhousing (Wilkinson et al., 1982) have also been shown toaffect louse populations.

Despite the ubiquity of lice, there have been very fewlongitudinal studies of natural louse infestations since the1940s; most studies have been relatively short-term com-parisons of louse prevalence and intensity, particularly incattle and sheep. However, long-term studies are valuablebecause they allow a more detailed understanding of thefactors which predispose individual animals to pediculosisand allow the effect of change in these factors over time tobe identified. The aim of the work described in this paper,therefore, was to undertake a longitudinal study of louseinfestation in four populations of donkeys over a period of21 months. The lice infesting these animals had previouslybeen shown to demonstrate high levels of pyrethroid toler-ance (Ellse et al., 2012). This tolerance was thought to havedeveloped through routinely repeated blanket treatmentof all donkeys with pyrethroid formulations which did notadequately permeate the coat. The work described hereaimed to identify risk factors associated with high infesta-tion intensities which might contribute to the developmentof more effective louse management.

2. Materials and methods

2.1. Study populations

Four farms, with large, long-standing equine popu-lations were used for the study. Each farm containedbetween 252 and 517 animals, the majority of which weredonkeys although some ponies, mules and hinnies werealso present. All animals were housed in groups typicallycontaining between 20 and 120 animals. All animals ateach farm carried a unique identification collar (as well asbeing micro-chipped). The farms were all located near Sid-mouth in the UK; all were geographically separated and thetransfer of animals between farms was infrequent. Prior tointroduction onto a farm, animals were quarantined andtreated for lice. Hence, louse infestation transfer betweenfarms was minimal and the farms can be considered to berelatively independent populations.

The study was carried out over 21 months from May2011 to January 2013. Each month, 24 animals wereexamined at each farm; only the donkey population wassampled. The sample size was estimated based on anexpected 10% prevalence of pediculosis, detected with 90%


confidence (Win Episcope 2.0). These figures were con-servative estimates, based on preliminary observations.Animals for examination were selected using a randomnumber generator (Excel, 2007, Microsoft). Animals were

1 <5 lice seen in all 5 the partings2 5–10 lice seen in all 5 partings3 >10 lice in all 5 partings

identified by their collars; when collars where missinga microchip scanner was used. During the course of thestudy, an outbreak of contagious disease resulted in thetemporary unavailability of some of the farms, but at leastone farm was visited and the animals examined everymonth throughout the 21 month study.

2.2. Louse abundance and excoriation

At each examination, a detailed data-capture sheetwas used to record information about each animal. Thisincluded notes about each animal’s conformation, hus-bandry, groom opinion, as well as the intensity of louseinfestation and self-excoriation. Louse abundance was esti-mated at five body sites: the supraorbital fossa, behind theear, on the side of the neck, in the axilla of the front legand half-way along the midline. Preliminary studies hadindicated that these were the infestation sites where licewere most likely to be found. In each area, the hair wasparted five times in 5 cm lengths. The estimated abundanceof lice in the five partings was assigned a score (Table 1).Each score was recorded separately for each body site. Theabundance score for nymphs in each region was recordedseparately from the score for adults, but the three nymphalinstars were grouped together, as it was not possible to dis-tinguish easily between nymphal stages by eye. Eggs wereonly recorded if they were found within the first 2 cm ofhair, as lice position their eggs close to the skin of theirhost and therefore eggs near the tips of the hair follicle arelikely to have hatched or be non-viable (Murray, 1957).

Donkeys were examined for visual indications ofself-excoriation, consistent with hair disturbance andshortening of the hair fibre. The size of the rub area wasrecorded on animal line-diagrams. On each donkey, thesum of all rub patches was assigned an excoriation cov-erage score based on the size of the rubbed area(s) relativeto the size of the donkey (Table 2).

2.3. Animal husbandry


rubbing.

Percentage of the donkeycovered by rub lesion

0 <10% 10–20% >30%

Excoriation Coverage Score 0 1 2 3


ING ModelV

Parasito

uerclctr((GMmas

2

sIlasAftelapss

intotushcvttCsKa

3

3

lswtw

ARTICLEETPAR-7219; No. of Pages 7

L. Ellse et al. / Veterinary

nknown, an estimate was made by an equine dentist. Atach visit the housing and bedding type were observed andecorded. Grooms were asked when the animal was lastlipped. To ascertain whether the extent and intensity ofouse infestation was accurately estimated by staff in closeontact with the animals, grooms were asked whether theyhought each donkey had lice, whether they had observedubbing and whether they would describe the rubbing as:1) hardly ever (not a noticeable amount), (2) sometimes,3) a lot of the time or (3) so much that it is causing concern.rooms were allowed to decline to answer this question.ean monthly temperature, calculated as the average ofean daily maximum and mean daily minimum temper-

tures, was obtained from one of two weather stations,ituated within 10 km of all the farms sampled.

.4. Statistical analysis

Most statistical analyses were carried out using R (ver-ion 2.15.1, 2012, R Foundation for Statistical Computing,SBN 3-900051-07-0). The frequency distribution of totalouse abundance scores followed a negative binomial, with

large number of zeros. This skewed distribution per-isted even when non-infested individuals were removed.s these distributions were considered too over-dispersed

or most statistical models, the data were analysed usingwo separate generalised linear models: a binomial modelstimating the proportion of the population with pedicu-osis using presence/absence (infested/not-infested) datand secondly a negative binomial model (GLM.nb, MASSackage, Venebles and Ripley, 2002) examining the inten-ity of infestation in louse-infested individuals, by using theummed total louse abundance score data.

In the multifactor analyses, interaction terms were notncluded as the multiplicity of interactions produced by theumber of explanatory factors in the analysis was likelyo induce Type 1 (false positive) errors. Stepwise removalf non-significant terms was carried out primarily usinghe step function and then further simplified using manualpdate commands to produce the best minimal model. Chi-quared was used to determine whether removal of a factorad a significant effect on residual deviance, if this was thease the factor was retained. Each continuous explanatoryariable was checked for linearity using generalised addi-ive model (GAM) plots. Non-linear functions and breaks inhe data were applied where appropriate. Finally, SPSS (IBMorp. Released 2010. IBM SPSS Statistics for Windows, Ver-ion 19.0. Armonk, NY: IBM Corp.) was used to calculate theappa statistic for the agreement between groom opinionnd researcher observation of louse infestation.

. Results

.1. Seasonality

All the lice observed were Bovicola (Werneckiella) ocel-atus; no other species were detected. There was a strong


easonal pattern, with the highest number of donkeysith lice and the highest louse abundances recorded in

he winter months. This generally coincided with periodshen donkeys were housed indoors and when the mean

PRESSlogy xxx (2014) xxx–xxx 3

monthly temperature was relatively low (Fig. 1). Basedon inspection of GAM plots the relationships betweenlouse abundance and temperature were split in both bino-mial and negative binomial models at 12.8 ◦C and 7 ◦Crespectively. Lower temperatures were associated withsignificantly higher numbers of donkeys with lice; thiswas statistically significant at temperatures both below(Z = −7.32, df = 1428, P < 0.05) and above 12.8 ◦C (Z = −7.57,df = 1428, P < 0.0.5). Total louse abundance scores were alsohigher in the cooler months, and this was significant attemperatures both below (Z = −3.35, df = 657, P < 0.05) andabove 7 ◦C (Z = −3.76, df = 657, P < 0.0.5). More donkeys hadlice when they were housed indoors (Z = −2.93, df = 737,P < 0.05) and the louse abundance scores were also higherduring this period (Z = −2.93, df = 737, P < 0.05). However,due to the strong correlation between temperature andhousing (t = −48.4, df = 1276.8, P < 0.05) removal of hous-ing from the multi-factor model did not have a significanteffect in the binomial analysis (P = 0.12) or louse abundancescore analysis (P = 0.76).

3.2. Coat distribution and predisposing factors

Of the sites inspected, more donkeys had lice in theaxillae than in any other body region (Z = 18.6, df = 562,P < 0.05; Fig. 2a). Of donkeys with lice the mean louse abun-dance score was also highest in the axillae (Z = 12.7, df = 879,P < 0.05; Fig. 2b). The axilla was also the region with thelongest hair length (Z = 28.4, df = 1561, P < 0.05; Fig. 2). Thepattern of infestation was similar for each life-cycle stage,with the axilla being the most commonly infested area.However nymphs and eggs were more commonly foundin the supraorbital fossa and along the midline than adultlice (Fig. 2).

For analysis of the overall effect of hair length onlouse burden, the mean of the five hair length measure-ments was calculated for each donkey. Based on inspectionof GAM plots the hair length data was split at 4.2 cm.More donkeys with longer hair had lice; this was statis-tically significant in groups with mean hair lengths below(Z = 4.43, df = 411, P < 0.05) and greater than 4.2 cm (Z = 4.26,df = 1411, P < 0.05). In donkeys with lice, there was no sig-nificant effect of hair length on their total louse abundancescore (Z = 1.64, df = 660, P = 0.1).

The relationship between donkey age and louse pres-ence was non-linear, with the best fit approximating aconvex quadratic curve meaning that; significantly morevery young and very old donkeys had lice than middle-ageddonkeys (Z = 4.83, df = 1441, P < 0.05) and these animals alsodisplayed higher louse abundance scores (Z = 3.27, df = 660,P < 0.05).

3.3. Self-excoriation

Donkeys with a higher excoriation coverage score weremore likely to have lice (Fig. 3). More than 80% of ani-mals which had an excoriation coverage score of three had


lice, this was statistically higher than animals which dis-played no signs of excoriation (Z = 2.78, df = 1411, P < 0.05;Fig. 3). Donkeys with higher excoriation coverage scoresalso had higher total louse abundance scores (Z = 34.2,


ARTICLE IN PRESSG ModelVETPAR-7219; No. of Pages 7

4 L. Ellse et al. / Veterinary Parasitology xxx (2014) xxx–xxx

0

2

4

6

8

10

12

14

16

18

20

0

10

20

30

40

50

60

70

80

90

Tem

pera

ture

( °C)

% o

f d

on

keys w

ith

lic

e

0

2

4

6

8

10

12

14

16

18

20

0

2

4

6

8

10

Te

mp

era

ture

(° C)

M

ea

n t

ota

l lo

use

ab

un

dan

ce

sc

ore

A

B

mean tated by

Fig. 1. The percentage (±se) of examined donkeys with lice (A) and the(bars). The periods when donkeys had access to outdoor grazing are indicclosed bars. Also, the mean monthly temperature (line).

df = 664, P < 0.05; Fig. 3). Louse abundance scores were sta-tistically higher in animals which had excoriation coveragescores greater than two, than animals which displayed noexcoriation (Z = 2.26, df = 659, P < 0.05; Fig. 3).

3.4. Louse management

Fewer donkeys that had been clipped 2–12 weeks ago


had lice than those that had not been clipped (Z = −2.2,df = 1411, P < 0.05). This pattern was seen when donkeyswere housed outdoors (Z = −5.06, df = 1043, P < 0.05; Fig. 4),but not when animals were housed indoors (Z = 0.82,

otal louse abundance score (±se) (B) of donkeys examined each monthopen bars and the periods when they were kept indoors are indicated by

df = 492, P = 0.41; Fig. 4). There was no significant effect ofclipping on the total louse abundance scores of donkeyswith lice, two weeks or more after clipping (Z = −1.44,df = 724, P = 0.15).

The grooms correctly identified whether an animalhad lice or not 63% of the time. The grooms underesti-mated the number of donkeys with lice; the sensitivityof their answers was 37.5% and their specificity was


87.3%. A Kappa coefficient of 0.258 indicates only aslight/fair level of agreement between researcher obser-vation and groom opinion. If grooms answered that theyhad not seen the animal rubbing, they were more likely to


ARTICLE IN PRESSG ModelVETPAR-7219; No. of Pages 7

L. Ellse et al. / Veterinary Parasitology xxx (2014) xxx–xxx 5

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0

5

10

15

20

25

30

35

40

Eye Ear Ne ck Axilla Mi dline

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Eye Ear Ne ck Axilla Mi dline

Mea

n h

air len

gt h

(cm)

Mea

n h

air len

gth

(cm)

% o

f d

on

key

s w

ith

lic

e

Supraorbital fossa

A

B

Supraorbital fossa

Infestation s ite

Infestation s ite

Mea

n l

ou

se a

bu

nd

an

ce s

core

Fig. 2. The percentage (±se) of donkeys examined (A) and the mean louseabs

ptopivsb(tatims

4

l

0

10

20

30

40

50

60

70

80

90

0

1

2

3

4

5

6

7

8

0 1 2 3

Perc

en

tag

e o

f do

nk

eys w

ith lic

e

Me

an

to

tal lo

us

e a

bu

nd

an

ce s

co

re

Excoriation cov erage score

not necessarily associated with a higher louse abundance,which may be regulated by different factors. For example,large differences in individual susceptibility to B. ovis have

0

20

40

60

80

100

120

Not 1/ 2w 2/ 4w 4w+

% o

f d

on

ke

ys

wit

h lic

e

Time si nce c lipped (weeks)

bundance score (±se) (B) of donkeys that were infested with, eggs (whitears), nymphs (grey bars) and adult lice (black bars) in each infestationite (bars). Also, the mean hair length of each infestation site (line).

redict that the animal did not have lice; their predic-ions of which donkeys had lice agreed with researcherbservations 61.7% of the time, but the sensitivity of theirrediction was just 18.8%. The Kappa coefficient of 0.143

ndicates only slight agreement between researcher obser-ations and groom predictions. If a donkey was seen to rubometimes by the groom there was a better agreementetween the researcher observation and groom opinion73.9% correct; Kappa = 0.335) although the grooms overes-imated the number of donkeys with lice in this group, with

specificity of 42.9%. If the donkey was seen to rub a lot ofhe time or more, then the number of donkeys that groomsdentified as having lice was overestimated, with no agree-

ent with researcher observations (Kappa = −0.056), highensitivity (95.8%) but no specificity.


. Discussion

The over-dispersed, negative binomial distribution ofouse prevalence, indicates a non-random distribution

Fig. 3. The mean total louse abundance score (±se) (bars) and the percent-age of donkeys inspected which had lice (±se) (line) in animals displayingdifferent excoriation coverage scores.

within the donkey population studied, suggesting thatsome animals are disproportionately susceptible to infes-tation (Tanner et al., 1980). However, notably, while thedata suggest that some factors, such as hair length, maybe associated with the presence of infestation they are


Fig. 4. The percentage of donkeys inspected which had lice (±se) whenhoused outdoors (white bars) or indoors (grey bars) and that have beenclipped: 1–2 weeks ago (1/2w), 2–4 weeks ago (2/4w), 4–12 weeks ago(4w+) or not within the last 12 weeks (Not).


ING Model

Parasito

ARTICLEVETPAR-7219; No. of Pages 7

6 L. Ellse et al. / Veterinary

been reported in sheep, with some experimentally infestedanimals failing to obtain a clinical infestation even afterprolonged housing with other infested sheep and repeatedintroduction of live lice into the fleece (James, 1998). Dif-ferences have been attributed to variations in immuneresponse; skin surface antibody titres for B. ovis were foundto be negatively correlated with the abundance of lice onewes in declining louse populations (James et al., 1998).Here, very old and very young donkeys were more likelyto carry a higher burden of lice than middle-aged animals.Since these two groups are likely to represent naive andimmunocompromised individuals respectively, the datasupport the idea that an immune-mediated response playsa role in regulating louse abundance. Hence, it is importantto distinguish between the causes of low levels of infes-tation, which are likely to be almost ubiquitous in manyanimal populations, and clinically significant infestations.

Predilection for certain areas of the coat has been notedin B. bovis infesting cattle (Devaney et al., 1988; Matthysse,1946) and here lice were more likely to be most abundantin the axillae. This aggregated distribution on the host maybe due to the protection from grooming and the more suit-able microclimate some sites provide (Devaney et al., 1988;Matthysse, 1946).

In the population of donkeys sampled, pediculosisdisplayed a strong seasonal pattern, with lower preva-lence and lighter burdens of lice in the summer months(April–September) and higher prevalence and heavier bur-dens in the winter months (October–March) (Fig. 1). Thisis consistent with findings in sheep and cattle in whichsimilar seasonality has been recorded (Craufurd-Benson,1941; James, 1998; Matthysse, 1946; Murray and Gordon,1969). The seasonal patterns correspond with changes inmean monthly ambient temperature, but the negative rela-tionships between louse abundance and mean ambienttemperature were non-linear. There was a breakpoint inthe binomial data at 12.8 ◦C suggesting that there is athreshold temperature, below which lice are more preva-lent within the population. This temperature may alsocorrespond to changes in management and behaviour ofthe animals leading to an increased rate of transmission.A similar breakpoint for louse abundance was seen but ata slightly lower temperature of 7 ◦C, suggesting that lousepopulations grow most rapidly at very low ambient tem-peratures. It cannot be determined however, whether thisis because lice survive and reproduce better in winter orsimply that they fare worse in summer. For example, stud-ies in sheep and horses, found that solar radiation duringthe summer could result in skin temperatures in excessof 45 ◦C in some regions of the coat (Murray, 1968, 1963)and that at this temperature B. equi adults and eggs cannotsurvive after 2 h of exposure (Murray, 1963).

Cooler temperatures are often associated with changesin management practices, such as moving animals frompasture to indoor housing, as is the case with the popula-tions of donkeys studied. Here a greater proportion of thepopulation were infested and higher louse burdens were


seen when animals were housed indoors, but this relation-ship was rendered non-significant by the much strongereffect of temperature in the multi-factorial model. Never-theless, from a practical perspective, housing may still be

PRESSlogy xxx (2014) xxx–xxx

a valuable indicator in that donkey carers should expect ahigher level of infestation when it is cold enough to houseanimals indoors.

The impact of infestation with chewing lice in cattle andsheep has been studied in conjunction with the effects ofexcoriation on leather and wool quality (Heath et al., 1995;Wilkinson et al., 1982). However, there has been limitedresearch on the visible manifestations of natural popula-tions of equine chewing lice as a result of self-excoriation.This study shows that there is a relationship between lev-els of visible coat rubbing and louse infestation in donkeys.The amount of rubbing a donkey displayed was positivelycorrelated with the presence of lice; 80% of donkeys withan excoriation score of three had lice, compared with just35% of donkeys that had no rubbed patches. These find-ings demonstrate that pediculosis has a negative impacton equine welfare and that hair fibre shortening and lightrubbing are good indicators of louse infestation. This sup-ports findings in sheep in which 32 naturally infested ewesshowed significant correlation between fleece derange-ment scores and burdens of B. ovis for 37 weeks (Jameset al., 2007).

Traditionally, shearing or clipping of animals has beenassociated with a drop in louse numbers and has beensuggested as a pediculosis control measure in cattle(Allen and Dicke, 1952; Murray, 1968; Wilkinson et al.,1982). Understanding the efficacy of clipping on lousenumbers is difficult because, on the farms sampled, ani-mals that appeared to be suffering from discomfort werealso often clipped and bathed with medicated shampoo(DermolineTM, Battles, Hayward & Bower Ltd.) to soothe theirritation and, unsurprisingly therefore, statistically moredonkeys that had been clipped within the last two weekshad lice than animals that had not been clipped. However,donkeys that were clipped more than two weeks ago hadfewer lice than unclipped individuals in the summer butnot in winter. This indicates that clipping was effectivein reducing the abundance of lice over time, but only inthe summer months when animals were outdoors whenhigher ambient temperature or solar radiation may haveresulted in higher louse mortality. It also likely that anynegative impact of clipping on the microclimate of the coatwas minimised during the winter months as donkeys wererugged to protect them from the cold. Overall these datasuggest that clipping donkeys is not an effective meansof louse control when animals are in winter housing. Thiscontradicts findings in cattle in which clipping at the peakof infestation was found to be an effective means of lousecontrol for up to 4 months (Allen and Dicke, 1952).

Louse prevalence was generally underestimated by thegrooms. In this population of donkeys, since the failure ofroutine pyrethroid treatments, the management of louseinfestation has been by targeted selective treatment ofwhat are assumed to be the most heavily infested ani-mals, based on observations of rubbing. Therefore, thislevel of underestimation is likely to result in poor levelsof louse control. If grooms had seen a donkey rubbing they


were more likely to answer that they thought the animalhad lice. However, excoriation and dermal discomfort mayalso be associated with a range of other skin conditionsin donkeys, such as dermatophytosis, dermatophilosis and


ING ModelV

Parasito

mseblbi

slaalimsrtodtrecu

tmyhpcshmFthtoascma

ARTICLEETPAR-7219; No. of Pages 7

L. Ellse et al. / Veterinary

ange (Svendsen, 2008). Therefore, while this study hashown that a correlation between louse prevalence andxcoriation does exist, the apparent overreliance on rub-ing behaviour as an indicator of louse abundance may

ead to misdiagnosis. If targeted selective treatment is toe used, a more robust protocol for the detection of louse

nfestation is needed.Based on the findings in this study, an effective control

trategy may be to mechanically control the louse popu-ation by clipping all animals during the summer monthsnd then to treat with insecticide in late autumn whennimals are brought in to winter housing and louse popu-ations begin to increase. Whilst whole heard treatments generally considered to be a poor resistance manage-

ent technique, a once yearly treatment is unlikely toelect for resistance. It is also likely to limit the need forepeated re-treatment of many individuals over the win-er period as animals are kept in closed herds. Rotationf drug classes between yearly treatments would furtheriminish selection pressure. However, it should be notedhat the scope for drug class rotation in this population isestricted as these lice have previously displayed high lev-ls of pyrethroid tolerance (Ellse et al., 2012) and there areurrently no other insecticide classes registered for equinese. Therefore, alternative pediculicides are required.

This study has shown that there is a strong seasonal pat-ern of donkey pediculosis which is correlated with mean

onthly ambient temperature. Also, that very old and veryoung donkeys are more susceptible to lice and that longerair predisposes animals to infestation and may afford licerotection from temperature associated adverse summerlimate. Clipping donkeys may provide a suitable treatmenttrategy in the summer; however, it does not appear toave an effect on louse numbers in the winter when ani-als are housed and louse burdens are at their highest.

or detection of lice, excoriation leading to shortening ofhe hair fibres in the coat is indicative of louse infestation;owever, over-reliance on rubbing behaviour as an indica-or of louse burden is likely to lead to incorrect estimationsf the burden of lice in the population. Therefore, without

sensitive and rigorous louse detection regimen, targeted


elective treatment of lice will not provide adequate louseontrol suggesting that summer clipping and a yearly treat-ent of animals prior to winter housing may be the most

ppropriate control measure in this population of donkeys.

PRESSlogy xxx (2014) xxx–xxx 7

Acknowledgements

We are grateful to the Donkey Sanctuary for funding thiswork and to Nikki Stradling and Krisin Hayday for their helphandling donkeys. The studies described were approved bythe University of Bristol ethics committee.

References

Allen, N.N., Dicke, R.J., 1952. Cattle lice control by clipping. J. Econ. Entomol.45, 970–973.

Craufurd-Benson, H.J., 1941. The cattle lice of Great Britain. Part 1: Biology,with special reference of Haematopinus eurysternus. Parasitology 33,331–342.

Devaney, J.A., Rowe, L.D., Craig, T.M., 1988. Density and distribution ofthree species of lice on calves in central Texas. Southwest. Entomol.13, 125–129.

Ellse, L., Burden, F., Wall, R., 2012. Pyrethroid tolerance in the chew-ing louse Bovicola (Werneckiella) ocellatus. Vet. Parasitol. 188,134–139.

Gawler, R., Coles, G.C., Stafford, K.A., 2005. Prevalence and distribution ofthe horse louse, Werneckiella equi equi, on hides collected at a horseabattoir in south-west England. Vet. Rec. 157, 419–420.

Heath, A.C.G., Cole, D.J.W., Bishop, D.M., Pfeffer, A., Cooper, S.M., Risdon,P., 1995. Preliminary investigations into the etiology and treatment ofcockle, a sheep pelt defect. Vet. Parasitol. 56, 239–254.

James, P., 1998. Seasonal dynamics and variation among sheep in den-sities of the sheep biting louse, Bovicola ovis. Int. J. Parasitol. 28,283–292.

James, P.J., Bartholomaeus, F.W., Karlsson, L.J.E., 2007. Temporal relation-ship between infestation with lice (Bovicola ovis Schrank) and thedevelopment of pruritic behaviour and fleece derangement in sheep.Vet. Parasitol. 149, 251–257.

James, P.J., Moon, R.D., Ragsdale, D.W., 1998. Serum and skin surfaceantibodies and their associations with sheep biting lice, Bovicolaovis, on experimentally infested sheep. Med. Vet. Entomol. 12,276–283.

Matthysse, J.G., 1946. Cattle lice. Their biology and control. Cornell Univ.Agric. Exp. Stn. Bull. 832, 3–97.

Murray, M., 1968. Ecology of lice on sheep. VI. The influence of shearingand solar radiation on populations and transmission of Damalinia ovis.Aust. J. Zool. 16, 725–738.

Murray, M.D., 1957. The distribution of the eggs of mammalian lice on theirhosts. I. Description of oviposition behaviour. Aust. J. Zool. 5, 13–18.

Murray, M.D., 1963. Influence of temperature on the reproduction ofDamalinia equi (Denny). Aust. J. Zool. 11, 183–189.

Murray, M.D., Gordon, G., 1969. Ecology of lice on sheep VII populationdynamics of Damalinia-ovis. Aust. J. Zool. 17, 179–186.

Svendsen, E., 2008. The Professional Handbook of the Donkey, 4th ed.Whittet Books.

Tanner, C.E., Curtis, M.A., Sole, T.D., Gyapay, K., 1980. The non-random,


negative binomial distribution of experimental Trichinellosis inrabbits. J. Parasitol. 66, 802–805.

Wilkinson, F.C., Dechaneet, G.C., Beetson, B.R., 1982. Growth of popula-tions of lice, Damalinia-ovis, on sheep and their effects on productionand processing performance of wool. Vet. Parasitol. 9, 243–252.


http://refhub.elsevier.com/S0304-4017(14)00221-0/sbref0005



























































































































































































































































































































































Documents

Seasonal infestation of donkeys by lice: Phenology, risk factors and management