It

Da

b

c

a

ARRA

KDSAP

1

i

0d

Veterinary Parasitology 173 (2010) 307–316

Contents lists available at ScienceDirect

Veterinary Parasitology

journa l homepage: www.e lsev ier .com/ locate /vetpar

n vitro and in vivo acaricidal activity and residual toxicity of spinosado the poultry red mite, Dermanyssus gallinae

.R. Georgea,∗, R.S. Shiel a, W.G.C. Applebyb, A. Knoxc, J.H. Guya

School of Agriculture, Food and Rural Development, Newcastle University, Newcastle upon Tyne, NE1 7RU, UKElanco Animal Health, Lilly House, Basingstoke, Hampshire, RG24 9NL, UKRoslin Nutrition Ltd., Gosford Estate, Aberlady, Longniddry, East Lothian, EH32 0PX, UK

r t i c l e i n f o

rticle history:eceived 7 April 2010eceived in revised form 21 June 2010ccepted 24 June 2010

eywords:ermanyssus gallinaepinosadcaricideoultry red mite

a b s t r a c t

This paper describes two experiments conducted to examine the acaricidal potential ofspinosad against the poultry red mite, Dermanyssus gallinae (De Geer), a serious ectopara-sitic pest of laying hens. Spinosad is a natural product derived from the fermentation of themicro-organism Saccharopolyspora spinosa. In vitro testing confirmed that, when appliedto a galvanised metal plate to the point of run-off, spinosad was toxic to adult female D.gallinae and suggested that at an application rate of 3.88 g/L a significant residual toxic-ity of spinosad could be achieved for up to 21 days. A subsequent in vivo experiment in aconventional cage housing system for laying hens demonstrated the acaricidal activity andresidual toxicity to D. gallinae of a single application of spinosad when applied at either1.94 or 3.88 g/L. Residual toxicity of spinosad at both of these application rates was main-tained throughout the course of the 28 day post-spray study period, with a peak in productefficacy seen 14 days after spraying. The results suggest that the greater the D. gallinae pop-ulation the greater will be the toxic effect of spinosad. Although the exact reasons for thisare unclear, it can be speculated that conspecifics spread the product between each othermore efficiently at higher mite population densities. However, further study is warrantedto confirm this possibility. Application of spinosad in vivo had no effect on hen bodyweightor egg production parameters (number and weight), suggesting that this product could beused to effectively control D. gallinae infestations whilst birds are in lay. This paper alsodescribes a novel method for effectively and efficiently achieving replication of treatments

in a single poultry house, previously unpopulated with D. gallinae. Individual groups ofconventional cages were stocked with hens, seeded with D. gallinae and used as replicates.Independence of replicates was achieved by isolating cage groups from one another usinga non-drying glue barrier to minimise D. gallinae migration. Creating isolated populations(replicates) of D. gallinae within a single poultry house thus represents a novel and efficientmeans of screening other potential acaricides under field conditions.

. Introduction

The poultry red mite, Dermanyssus gallinae (De Geer),s currently the most economically deleterious ectopara-

∗ Corresponding author. Tel.: +44 0 1524510900; fax: +44 0 152461536.E-mail address: d.george@lancaster.ac.uk (D.R. George).

304-4017/$ – see front matter © 2010 Elsevier B.V. All rights reserved.oi:10.1016/j.vetpar.2010.06.035

© 2010 Elsevier B.V. All rights reserved.

site of laying hens in Europe (Chauve, 1998). Worldwide, D.gallinae prevalence in laying flocks varies from 20% to 90%,depending upon the country and production system con-

sidered (Sparagano et al., 2009). In the UK, infestation levelsof between 60% (Fiddes et al., 2005) and 85% (Guy et al.,2004) can be expected in commercial egg laying premises,with higher mite populations typically seen in free-rangesystems compared to cage units (Guy et al., 2004; Fiddes

y Parasi

308 D.R. George et al. / Veterinar

et al., 2005; Arkle et al., 2006). This is of particular concerngiven that conventional cages will be prohibited in the EUfrom 2012 (EU Directive 99/74/EC) and thus the proportionof hens housed in alternative systems such as free-range islikely to increase substantially.

Where infestations of D. gallinae are sufficiently severe,mite feeding may result in significant stress to hens witha resulting negative impact on bird condition, growth rate,egg quality (through increased shell thinning and spotting)and egg production (Urquhart et al., 1996; Chauve, 1998).In extreme cases, D. gallinae population levels may be sohigh that anaemia, and even death of hens, can result frominfestation (Wojcik et al., 2000; Cosoroaba, 2001). Whilstlow level mite infestation may pose fewer problems to thehost birds, it has been reported that D. gallinae may serveas a vector for numerous poultry pathogens (Chirico et al.,2003; Valiente Moro et al., 2009). This suggests that evensmall numbers of D. gallinae in poultry systems may have aserious impact upon the production and welfare of layinghens.

Economic costs associated with D. gallinae due to bothcontrol and reduced production have been estimated atD130 million annually for the EU egg industry (van Emous,2005). Control of D. gallinae has typically been achieved bythe use of synthetic contact acaricides including carbaryl,diazinon, dichlorvos and permethrin, but the continued useof these products is hampered by issues of mite resistance(Beugnet et al., 1997; Kim et al., 2004, Fiddes et al., 2005)and decreasing product availability as a consequence ofmore stringent legislation.

Spinosad was first registered for agricul-tural/horticultural use in the late 1990s and by 1999it was approved for use on over 100 crops in 24 differentcountries (Thompson et al., 2000). Being a natural productderived from fermentation of the micro-organism Saccha-ropolyspora spinosa (Mertz and Yao), spinosad possessesseveral favourable characteristics for a pesticide. Forexample, Anastas et al. (1999) report that spinosad willnot bio-accumulate, volatize or persist in the environmentand will degrade naturally when exposed to light. Inaddition, spinosad has been found to display activityagainst a range of insect pests, especially those in thegenera Lepidoptera, Diptera and Thysanoptera, and to alesser extent the Coleoptera and Orthoptera (Thompsonet al., 2000). Corresponding research has also found thatspinosad possesses relatively low toxicity to mammalsand birds, where acute oral LD50 values of >5000 and>2000 mg/kg have been reported for mice and mallardducks, respectively (Thompson et al., 2000). Researchalso suggests that 70–90% of beneficial insects are leftrelatively unharmed by spinosad (Anastas et al., 1999),where for certain ladybirds, lacewings and predatorymites, LD50 values are >1000× greater than with cyper-methrin (Thompson et al., 2000). The toxicity of spinosadto mites per se has been reported as being variable and/orreduced in comparison to insect species (Thompson et al.,

2000; Villanueva and Walgenbach, 2006; Holt et al., 2006).Experiments using spinosad against the ectoparasiticcattle tick Boophilus microplus have nevertheless yieldedpromising results (Davey et al., 2001, 2005), as has workwith other soft and hard tick species (Cetin et al., 2009).

tology 173 (2010) 307–316

It is therefore possible that spinosad could provide a newand effective control product for use against D. gallinae,whilst fulfilling additional desirable environmental andnon-target organism toxicity criteria.

Therefore, the aim of this study was to test the acaricidalactivity of spinosad to D. gallinae at different concentrationsand varying times after product application. This was doneinitially using an in vitro design in a laboratory experiment,followed by an in vivo experiment to determine the effec-tiveness of spinosad against D. gallinae under conditionswhich mimicked commercial egg production.

2. Materials and methods

2.1. In vitro experiment

2.1.1. Experimental design and treatmentsThe experiment was designed as a four × eight fac-

torial, with four application rates of spinosad and eighttime points post-spraying (PS) giving 32 exposure treat-ments in total. Spinosad solutions were made using thecommercial product Elector® (Elanco Animal Health, Bas-ingstoke; 452.8 g/L spinosad) diluted in distilled waterto give concentrations of 0.00, 0.97, 1.94 and 3.88 g/Lspinosad. Solutions were applied to the test surfaces asa coarse spray using a standard 500 ml hand-operatedatomiser. Surfaces were sprayed to the point of run-off and allowed to air dry. For each application rate ofspinosad, D. gallinae were exposed to the treated sur-faces at one of eight time points PS; 2.5 h, 1, 3, 7, 10,14, 21 and 28 days. There were four replicate test unitsper treatment, and the study was implemented by atechnician who remained blinded to the experimentaltreatments.

2.1.2. Source of D. gallinae and test apparatusD. gallinae were collected from a free-range laying

unit in Northumberland (England), brought to the labo-ratory at Newcastle University (England) and stored fora period of 48 h prior to use. Recently fed adult femalemites were then placed in the test unit, namely a galvanisedmetal plate (70 mm × 70 mm × 3 mm; Metal Supermarket,Gateshead, England) which had been previously coatedwith one of the four different application rates of spinosad.Mites were contained on the metal plate by an invertedPetri dish, coated with the same concentration of spinosad.A Vaseline seal was used around the Petri dish base toensure that mites could not leave the test arena. The metalplates and Petri dishes were contained within a fumecupboard prior to the addition of D. gallinae, where condi-tions of continuous air-flow were maintained throughoutthe study in an attempt to mimic conditions in a ven-tilated poultry house. Once mites were placed into thetest unit, they were transferred to a climate-controlledgrowth room maintained at 22 ◦C with a 16:8 light:darkcycle.

Approximately 25 D. gallinae were used per test unit.The proportion of live (active, clear locomotion), moribund(passive, hardly any locomotion, but showing some move-ment, e.g. of legs) and dead D. gallinae (no movement)was calculated after exposure to the treatments for 48 h

D.R. George et al. / Veterinary Parasitology 173 (2010) 307–316 309

Fig. 1. Schematic representation of the experimental set-up used inside a poultry house to test the in vivo acaricidal activity of spinosad against Dermanyssusg rimentw bered (t es. Oneb

(t

2

‘tyMsitrfT

2

2

coauScceEp5dcb

t

allinae. Cages shaded in light grey represent those not used in the expeith hens and comprising a test unit, where these ‘cage groups’ were num

he poultry house for identification. Un-shaded cages represent buffer zonlocks shown.

±45 min). Mite status was assessed under 4× magnifica-ion by agitation with an entomological pin.

.1.3. Statistical analysisMoribund and dead mites were grouped together as

deceased’ for analysis of percentage mortality betweenreatments. The results were subjected to a two-way anal-sis of variance (ANOVA) using the statistical packageinitab (v 15; Minitab Inc., State College, USA) which con-

idered both application rate and time PS as factors. Thenteraction between these factors was also considered inhe analysis. One data point for 3.88 g/L 10 days PS wasemoved prior to analysis as a rank outsider, being 4.97 SDrom the mean. Post hoc testing was performed using theukey’s test.

.2. In vivo experiment

.2.1. Experimental set-upThe experiment was a randomised block design and

onsisted of three treatments, namely a control (waternly) and two application rates of spinosad (low = 1.94 g/Lnd high = 3.88 g/L). A small commercial layer house sit-ated in East Lothian (Roslin Nutrition Ltd., Longniddry,cotland) was used for the experiment. The house had notontained poultry for a considerable period of time andontained two rows of conventional cages for laying hens,ach with four tiers of cages set back-to-back (see Fig. 1).ach row was 24 cages long, providing a total of 192 cageser row. Each cage contained four nipple drinkers and was00 mm wide, 460 mm deep and 460 mm high at the front

eclining to 400 mm high at the back. At the front of eachage there was a feed trough and an egg collection belt,oth of which ran the entire length of the row.

To achieve replication of treatments within the poul-ry house, cages were split into ‘groups’, where any group

whilst those shaded in dark grey represent a group of six cages stocked1–12) in ascending order in a clockwise pattern from the front and left ofreplicate of each treatment used was cited at random in each of the four

comprised one column of six back-to-back cages (the uppertier of cages was not used in the experiment as these cageswere difficult to access). A group of cages was left emptyat the front of the house after which every fourth group ofcages was stocked with birds (leaving two groups of cagesempty at the rear of the house) (see Fig. 1). Thus, therewere four replicates for each of the three treatments ineach of four blocks, where a replicate consisted of one ofthe groups of six stocked cages (see Fig. 1). All stocked cagescontained four hens of a commercial genotype (LohmannBrown) which were approximately 55 weeks old at thestart of the experiment and had been previously housedin the same four-bird groups in another cage facility. Birdswere fed ad libitum with a commercial layer mash. As livebirds were used and were being infected with D. gallinae theexperiment was conducted under Home Office License andlocal guidelines on animal welfare at Newcastle University.

2.2.2. Establishing D. gallinae populationsIn order to establish isolated known populations of D.

gallinae within the poultry house, which had been emptyof poultry for some time, each group of stocked cageswas seeded with D. gallinae collected from a free-rangelaying unit in Northumberland (England). Mites were col-lected and weighed into transport chambers (adapted from250 ml plastic screw-top beakers with a mesh lid to allowair movement; Fisher Scientific, Leicestershire, England),such that each chamber contained approximately 20,000D. gallinae (of mixed age and sex). Mite numbers in thesechambers were estimated by weight, where the averageweight of 20,000 D. gallinae (of mixed age and size) had

been predetermined.

D. gallinae were released into stocked cages some 72 hafter being collected by securing one transport chamberto the egg tray of each cage and severing the mesh lid. D.gallinae numbers in all stocked cages were then assessed

310 D.R. George et al. / Veterinary Parasitology 173 (2010) 307–316

Table 1Proportion of adult female D. gallinae dead or moribund after application of different rates of spinosad and after different periods post-spraying (PS). Meansnot sharing a common letter (of the same case or format where they may be compared to one another) are significantly different (P < 0.05). Means aredisplayed with ± standard errors; n = 16 for all time PS means, except 10 days PS where n = 15. n = 32 for all application rate means, except 3.88 g/L wheren = 31. n = 4 for all time PS × application rate means, except 10 days PS × 3.88 g/L where n = 3.

Time PS Spinosad application rate (g/L)

0.00 0.97 1.94 3.88 Mean

2.5 h 0.043 ± 0.017 ab 0.775 ± 0.061 ijk 0.868 ± 0.030 jk 0.933 ± 0.013 k 0.655 ± 0.094 a1 day 0.052 ± 0.026 abc 0.310 ± 0.041 efg 0.667 ± 0.070 ij 0.739 ± 0.032 ijk 0.442 ± 0.075 b3 days 0.060 ± 0.029 ab 0.349 ± 0.103 gh 0.576 ± 0.052 hi 0.696 ± 0.019 ij 0.420 ± 0.068 b7 days 0.010 ± 0.010 a 0.101 ± 0.053 abcdef 0.321 ± 0.054 fg 0.558 ± 0.013 hi 0.248 ± 0.057 cd10 days 0.041 ± 0.024 ab 0.201 ± 0.058 abcdefg 0.292 ± 0.073 defg 0.612 ± 0.055 i 0.265 ± 0.058 c14 days 0.063 ± 0.028 abc 0.128 ± 0.039 abcdefg 0.206 ± 0.073 abcdefg 0.284 ± 0.029 cdefg 0.170 ± 0.030 d

0.0660.076

0.384

21 days 0.032 ± 0.011 a 0.076 ± 0.036 abcde28 days 0.011 ± 0.011 a 0.124 ± 0.040 abcdefg

Mean 0.039 ± 0.007 A 0.258 ± 0.043 B

weekly during a 28 day mite establishment phase to ensurethat mite seeding had been successful. Estimates of D. gal-linae populations were conducted by trapping mites in aspecialised device (Mite Monitor, ADAS, Wolverhampton,England) which contained a removable mite trap. Each trapcomprised a 2 mm fluted polypropylene board that wasfolded in half once and then slotted into the rigid plasticmonitor. The traps were left in place for 24 h (±1 h) on eachsampling occasion. Monitors were permanently fixed to thefloor of each stocked cage (beneath the feed trough out ofreach of the hens) using plastic cable ties.

On any one sampling occasion, the traps from the sixmonitors in any stocked cage group were removed andplaced together into a screw-top beaker containing 70%ethanol as a killing agent. Traps were returned to the labo-ratory in these beakers where they were washed in distilledwater to dislodge any mites within them. D. gallinae num-bers per stocked cage group (replicate) were then assessedunder magnification, where both the total number of mitesand the proportion of that total which comprised adultfemales were determined.

Each cage also contained an additional mite monitor andtrap therein fitted to the cage floor under the feed trough.These were left in place for the duration of the study to actas a reservoir for D. gallinae and aid initial mite populationestablishment.

To reduce the risk of D. gallinae migrating from onestocked cage group (replicate) to another, a ‘buffer zone’was put in place that comprised the three groups of emptycages situated between groups of stocked cages. The centrecages of each buffer zone were treated with two strips ofInsect Barrier Glue (IBG; Agralan Ltd., Swindon, England)to minimise D. gallinae migration between stocked cages.IBG was also used around cage supports to isolate the toptier of cages from the remaining cages used and to isolateall cages from the floor of the poultry house.

To quantify any D. gallinae migration which might havetaken place between groups of stocked cages, a mite moni-tor was placed in each central buffer zone cage between the

two strips of IBG applied. Mite traps were placed in thesemonitors at the start of the study (the point of D. gallinaeseeding) and replaced thereafter on a weekly basis (exceptduring Weeks 4–5 where an extra collection was madeto coincide with mite monitoring in stocked cage groups).

± 0.031 abcd 0.274 ± 0.050 bcdefg 0.112 ± 0.029 de± 0.037 abcde 0.073 ± 0.011 abcd 0.071 ± 0.016 e

± 0.052 C 0.518 ± 0.051 D

Traps were processed as previously described, where totalD. gallinae counts were determined for each buffer zone.

2.2.3. Treatment applicationFollowing the 28 day D. gallinae establishment phase,

cages were subject to one of three treatments, namelya control (water only) or one of two application ratesof spinosad (low = 1.94 g/L and high = 3.88 g/L, formulatedusing Elector® as previously described). Product wasapplied to cages by a pressurized hand-held lance sprayerto the point of run-off. Treatments were applied to allaccessible surfaces of stocked cages, as well as those ofimmediately adjacent cages (since D. gallinae could havebeen resident in these cages despite the absence of birds).

2.2.4. D. gallinae population monitoring post-treatmentD. gallinae populations in all stocked cages were

assessed at 3, 7, 14, 21 and 28 days post-treatment usingthe same monitors, traps and 24 h trapping protocol asdescribed previously. Numbers of adult female D. gallinaeand nymphs/adult males in the trap samples were countedby a technician who was blinded to the allocation of treat-ment to cage groups in the building.

2.2.5. Hen response to treatmentRecords of any hen mortality were kept during the study

with post-mortem analysis undertaken on any birds thatdied. In addition, all birds were group weighed on entry tothe poultry house and again at the end of the study. Finally,egg production per stocked cage group (quantity and aver-age weight) was also monitored daily for the duration ofthe study.

2.2.6. Statistical analysisIn order to examine the effect of treatment on popula-

tions of D. gallinae response parameters were analysed ona per time point PS basis (i.e. individual time points wereanalysed separately). Total mite numbers were analysedby a two-way AVOVA considering treatment and block as

factors, where data for 7 days PS were square-root trans-formed prior to analysis to better fit the residuals from theANOVA to a normal distribution (checked by application ofthe Anderson–Darling test). D. gallinae numbers retrievedfrom traps in Week 4 of the mite establishment phase

D.R. George et al. / Veterinary Parasitology 173 (2010) 307–316 311

Fig. 2. The total number of Dermanyssus gallinae in traps retrieved from stocked cage groups prior to treatment with spinosad at 1, 2, 3 and 4 weeks afterseeding. Cage groups were numbered in ascending order in a clockwise pattern from the front and left of the poultry house. Data for ‘Week 2 – cage group3’ have been corrected by a factor of 6/5 to account for one of the six traps being missing in this cage group. Data for ‘Week 3 – cage group 4’ and ‘Week3 – cage group 11’ have been corrected by a factor of 6/3 to account for three missing traps in each cage group. Data labels depict subsequent treatmentassignment to cage groups where Con: control (0.00 g/L spinosad), SpL: spinosad low (1.94 g/L spinosad) and SpH: spinosad high (3.88 g/L spinosad).

Fig. 3. The proportion of adult female Dermanyssus gallinae in traps retrieved from stocked cage groups prior to treatment with spinosad at 1, 2, 3 and 4weeks after mite seeding. Cage groups were numbered in ascending order in a clockwise pattern from the front and left of the poultry house.

Fig. 4. Mean total number of Dermanyssus gallinae in traps retrieved from stocked cage groups under different treatments at 3, 7, 14, 21 and 28 dayspost-spraying (PS). Means not sharing a common letter are significantly different (P < 0.05) within any given day PS, where ns: no significant differencebetween treatments. Means are displayed with ± standard errors. n = 4 for all means.

312 D.R. George et al. / Veterinary Parasitology 173 (2010) 307–316

from stdifferenr all me

Fig. 5. Mean proportion of female Dermanyssus gallinae in traps retrievedpost-spraying (PS). Means not sharing a common letter are significantlybetween treatments. Means are displayed with ± standard errors. n = 4 fo

(i.e. directly before application of treatment), were consid-ered as a covariate in this analysis. The proportion of adultfemales in samples was also analysed by a two-way ANOVAto provide information on population structure. Data col-lected 7 days PS was again square-root transformed priorto analysis.

Analysis of hen bodyweight, egg number and egg weightbetween treatments post-spraying was conducted using atwo-way ANOVA as with other data. Pre-spray data wereconsidered as a covariate in the analysis. Cages containingdead hens were excluded from the analysis in all cases asit was found (for hen bodyweight) that a reduced number

of birds per cage significantly affected the data (P < 0.001).Having excluded data from individual cages with deadbirds, per cage group data were corrected to provide a valuefor six full cages (i.e. a complete cage group) prior to anal-ysis where necessary.

Table 2Numbers of D. gallinae trapped in buffer zones during Weeks 1–8 of the study, wdisplayed with ± standard errors. Buffer zone numbers depict locations between4). Cage groups were numbered (1–12) in ascending order in a clockwise pattern

Buffer zone Week of study period

1 2 3 4 4/5

1/2 45 2 8 18 62/3 9 0 2 0 13/4 20 5 2 0 34/5 4 0 0 3 05/6 2 0 6 2 57/8 2 2 1 1 28/9 13 0 2 4 39/10 79 0 1 2 810/11 4 0 0 0 211/12 23 3 0 1 1

Mean 20.10 ± 7.78 1.20 ± 0.55 2.20 ± 0.85 3.10 ± 1.71 3.1

Table 3Growth and production parameters of hens housed under different treatments afstandard errors; n = 4 for all means. Data for cages in which hens died have been

Spinosad application rate (g/L) Hen bodyweight (kg)

0.00 1.84 ± 0.011.94 1.84 ± 0.023.88 1.86 ± 0.02

ocked cage groups under different treatments at 3, 7, 14, 21 and 28 dayst (P < 0.05) within any given day PS, where ns: no significant differenceans.

As for the in vitro experiment, all analyses were con-ducted using Minitab v15.

3. Results

3.1. In vitro

Both time PS and spinosad application rate had a highlysignificant effect on D. gallinae mortality (P < 0.001 in bothcases). There was also a highly significant interactionbetween these two variables (P < 0.001). Tukey’s tests wereused to look for significant differences between pairs of

here treatments were imposed at the beginning of Week 5. Means arecage groups (e.g. Buffer zone 3/4 was situated between cage groups 3 andfrom the front and left of the poultry house.

5 6 7 8

8 18 17 564 0 1 01 1 0 13 0 0 17 6 48 273 1 0 00 1 0 17 14 15 40 1 0 31 0 1 0

0 ± 0.80 3.40 ± 0.96 4.20 ± 2.06 8.20 ± 4.88 9.30 ± 5.80

ter application of different rates of spinosad. Means are displayed with ±removed prior to formulating means for any given cage group.

Egg production (eggs/bird/day) Egg weight (g)

0.83 ± 0.0 66.9 ± 1.60.84 ± 0.0 64.7 ± 2.60.84 ± 0.0 67.3 ± 0.9

means and the outcome of these are provided in Table 1.Application of spinosad consistently resulted in higher

D. gallinae mortality than the control treatment early inthe experiment. There was a clear dose-response for resid-ual toxicity, where as spinosad application rate increased

y Parasi

tha

3

3

wFctatWwfta2

3o

btaatbAttaab9t

p1tfma

iP=7e2oFdippPr

D.R. George et al. / Veterinar

he mortality of D. gallinae mortality remained significantlyigher than the control for a longer period of time PS; 3, 10nd 21 days PS for 0.97, 1.94 and 3.88 g/L, respectively.

.2. In vivo

.2.1. Establishing isolated D. gallinae populationsD. gallinae were found in all stocked cage groups in all

eeks of the D. gallinae establishment phase (Weeks 1–4;ig. 2). The average number of mites recovered from eachage group was 168 in Week 1, with this figure fallingo 132 in Week 2 and then rising to 267 in Week 3 andgain to 367 in Week 4. There was a notable change inhe proportion of adult female D. gallinae in traps between

eeks 1 and 2 (Fig. 3) of the mite establishment phase,here the average proportion of females in samples rose

rom 14% to 30%. In Week 3 the proportion of females inraps from cage groups had fallen slightly to 26% on aver-ge with a further slight decline seen in Week 4 (average:2% females).

.2.2. D. gallinae population monitoring post applicationf treatment

Total D. gallinae numbers were significantly differentetween treatments at all points PS, with the excep-ion of 21 days PS (P < 0.001, <0.001, <0.01, =0.062nd <0.01, respectively, for data collected 3, 7, 14, 21nd 28 days PS). Block had no significant affect onhe data at any time point PS. Significant differencesetween pairs of treatment means are provided in Fig. 4.lthough there were no significant differences between

he two spinosad application rates, there was a consis-ent trend for lower D. gallinae numbers at the higherpplication rate. For the data shown in Fig. 4, percent-ge reductions seen relative to the control peaked foroth spinosad application rates at 14 days PS (87% and7% reductions at the lower and higher rate, respec-ively).

For 3 and 7 days PS data, there was a significant effect ofre-spray D. gallinae population (P < 0.05 in both cases). For4, 21 and 28 day PS data, there was no significant effect ofhe pre-spray mite population. The regression coefficientsor the covariates were all <1, indicating that spinosad was

ore effective against larger populations of D. gallinae thangainst smaller ones.

The proportion of adult females in traps was signif-cantly different between treatments at all time pointsS except for 3 and 21 days PS (P = 0.293, <0.05, <0.01,0.056 and <0.01, respectively, for data collected at 3,, 14, 21 and 28 days PS). The effect of block was gen-rally non-significant, with the only exception being for8 days PS (where P < 0.05). Differences between pairsf treatment means at different points PS are shown inig. 5, where the Tukey’s test was also able to detectifferences for data collected at 21 days PS. Differences

dentified were inconsistent across the different timeoints as a whole. There was an initial drop in the pro-ortion of females in spinosad-treated cages at 7 daysS, but at all subsequent time points PS this trend waseversed.

tology 173 (2010) 307–316 313

3.2.3. Minimising D. gallinae movement betweenreplicates

Numbers of D. gallinae retrieved from buffer zones overthe entire study period were variable in some weeks, withsome zones harbouring relatively large numbers of mitesin traps (Table 2). However, overall numbers of D. gallinaetrapped in the buffer zones were consistently low, partic-ularly in Weeks 2–6 inclusive.

3.2.4. Hen response to treatmentFive birds (1.7% of the total used) died or were

euthanaised during the course of the study, well below thelevel of 5% mortality considered the norm under commer-cial conditions. Four of the five deaths occurred prior to theapplication of treatment to the cages, whilst the fifth deathwas culling due to a leg injury. Post-mortem examinationwas conducted on all five birds and revealed no unusualcauses of death. No more than one bird died in any oneindividual cage during the course of the study.

Analysis of hen bodyweight, egg number and egg weight(Table 3) showed no significant effect of treatment or blockon any of the data. Hen bodyweight on entry to the poultryhouse had a significant impact on bodyweight at the end ofthe study (P < 0.01). Egg production (but not weight) pre-spraying had a significant impact on egg production post-spraying (P < 0.05).

4. Discussion

4.1. In vitro

The results of the in vitro laboratory experiment con-firmed that spinosad was toxic to D. gallinae, although inno instance did exposure to surfaces treated with spinosadresult in 100% mite mortality. Observations of D. gallinaebehaviour between the treated and control plates sug-gested that mortality may have been increased had a longerperiod of exposure been used, since ‘live’ mites in spinosad-treated plates were often less active than those in controls.The mode of action of spinosad is characterized by excita-tion of the nervous system leading to subsequent paralysisand death (Anastas et al., 1999; Thompson et al., 2000).Knock-down is therefore not immediate and death maytake several days, hence the reason for grouping deceasedand moribund mites together for statistical analysis. It ispossible then that less active D. gallinae in spinosad-treateddishes may have been displaying early symptoms of toxic-ity and could have been expected to die had the exposureperiod been increased past 48 h. The observed knock-downwith spinosad was nevertheless still favourable when com-pared to a range of alternative D. gallinae control methods,several of which could also be expected to suffer from alack of residual toxicity (George et al., 2010).

These results suggested that at higher concentrationsspinosad displays some residual toxicity to D. gallinae, with60% of mites (when corrected to take account of any nat-

ural mortality observed in the control treatment) beingkilled over a 48 h period following exposure to the high-est application rate 10 days PS. Some residual toxicity (incomparison to the control) was seen 21 days PS at the high-est application rate, although by this point mortality was

y Parasi

314 D.R. George et al. / Veterinar

reduced to less than 30% (again, when corrected to takeaccount of any natural mortality observed in the controltreatment).

4.2. In vivo

The results of trapping during the D. gallinae establish-ment phase demonstrated that cages had been successfullyseeded with populations of D. gallinae. It appeared that,based on consistently increasing mite numbers in trap sam-ples, D. gallinae populations had established themselves inthe cage groups by Weeks 3 and 4 after an initial drop inmite numbers in Week 2. This drop may have been causedby the population going through a ‘maturing’ phase in thefirst week of establishment (during which the death rateof mites exceeded the birth rate as juvenile mites maturedto reproductive status), as supported by an increased pro-portion of females in traps in Week 2. The slight decreasein female proportions in Weeks 3 and 4 suggests that hav-ing matured, ‘new’ females were then productive and hadstarted to contribute to younger generations, thus givingrise to increased D. gallinae populations overall. The resultsalso suggested that D. gallinae populations had not estab-lished uniformly across all cage groups, probably due tominor variations between cage groups with regard to anynumber of uncontrollable variables (exact initial seedinglevels, location, humidity, air-flow, temperature, etc.). Toaccount for this variation, D. gallinae numbers retrievedfrom traps in Week 4 of the mite establishment phase(i.e. directly before application of treatment), were consid-ered as a covariate when analysing total D. gallinae countsobtained post-spraying.

Seeding of premises with D. gallinae is rarely undertakenin this type of study, where work is often conducted in fullystocked poultry houses already infested with mites. How-ever, achieving replication in such studies poses a problemwhere typically a single poultry house is treated in itsentirety, thus providing only one true replicate (e.g. Keïtaet al., 2006; Meyer-Kühling et al., 2007a,b). Creating iso-lated populations (replicates) of D. gallinae within a singlepoultry house represents a novel and efficient means ofscreening new potential acaricides under field conditions.The buffer zones created in the current study to minimiseD. gallinae migration between stocked cage groups workedwell in most weeks. Relatively high D. gallinae numberswere recorded in buffer zones in Week 1 of the study,which probably reflected higher than normal D. gallinaemovement immediately after seeding and before mites hadsettled in refugia. Increases in average D. gallinae numbersin buffer zone traps were also observed towards the endof the study period, possibly as a result of some IBG stripsbeing compromised by a build-up of dust and debris by thistime. However, as buffer zone trap catches were generallynegligible (compared to overall mite numbers in stockedcage groups), any low level D. gallinae migration that didoccur would not have been expected to affect the results.

D. gallinae populations appeared to experience a decline at21 and 28 days PS, the reasons for which remain unknown.It is possible that natural population fluctuation causedthe variation observed, where such fluctuation around anequilibrium population has been observed repeatedly for

tology 173 (2010) 307–316

D. gallinae in laying systems (Nordenfors and Höglund,2000; Arkle et al., 2004). However, changes in a number ofunrecorded external variables could also have caused thedecline seen, though it deserves note that the temperatureinside the house was maintained at a relatively constant20 ◦C throughout the study period.

In vivo results confirmed those of the in vitro experimentin finding that spinosad was toxic to D. galliane, where max-imum mite reductions of 97% were recorded at the highestapplication rate (3.88 g/L, 14 days PS). It may be possible toachieve more complete control of D. gallinae in the field byrepeat application of product at the concentrations used.Experiments using spinosad against other ectoparasiteshave produced similar conclusions, where repeat applica-tions have been recommended for effective tick control(Davey et al., 2001, 2005). Due to the reclusive life cycleof D. gallinae, repeat application of contact acaricides isoften recommended to ensure that the generation emerg-ing from hard-to-treat refugia post-treatment is targetedalong with any existing nymphs and adults. Bye Mite®

(50% phoxim) has been shown to be display 97–99% effi-cacy against D. gallinae when application at 2000 ppm wasrepeated after 7 days (Keïta et al., 2006; Meyer-Kühling etal., 2007a). Efficacies between 90% and 99% have also beenachieved with this product using the same spraying regime,but in ‘alternative’ housing systems (Meyer-Kühling et al.,2007b). Repeat application of spinosad would thus be likelyto increase efficacy, and be in keeping with existing man-agement practices for other approved products.

Interestingly, the results from the ANCOVAs conductedproduced regression coefficients for covariates that werealways less than 1, implying that spinosad had a greatereffect on larger D. gallinae populations. Reasons for thisremain unclear, but it is possible that at higher D. gallinaepopulation densities individual mites were exposed to agreater amount of product due to movement and spread ofthe active ingredient by conspecifics, where as previouslynoted spinosad does not kill target pests immediately (thuspermitting such spread). This type of ‘biological dispersal’may be a factor in the transmission of entomopathogenicfungi (Meyling and Eilenberg, 2007), including by D. galli-nae (Kilpinen et al., 2002), although it is rarely considered asa factor effecting the efficacy of other pest control products.However, further study would be required to determinewhether this speculation was true for spinosad in partic-ular. Such investigations could also provide insight intotreatment threshold levels and product application tech-niques.

In addition to displaying good toxicity to D. gallinae, theresidual field efficacy of spinosad in the current study alsoappeared favourable, where reductions in mite numbers of92% and 88% were seen 21 and 28 days PS, respectively (atthe highest application rate). Such reductions were unex-pected in light of the shorter residual toxicities displayedin the in vitro experiment, although this may reflect therelatively short (48 h) exposure period used in these tests.

Maximum reductions in D. gallinae populations (relativeto control populations) were achieved 14 days PS for bothspinosad treatments. That such a maximum effect wasachieved 2 weeks after spraying was not unexpected. SinceD. gallinae may not feed for several days, and develop from

y Parasi

tttsaDmtTtb

mfotwinNiwt

5

teawrsDobDcewbps

A

NRp

R

A

A

A

D.R. George et al. / Veterinar

he egg stage in relatively protected refugia, it was possiblehat some mites did not emerge from the relative protec-ion of their refugia until into the second week PS. This isupported by the fact that at 7 and 14 days PS there wassignificant reduction in the proportion of adult female. gallinae recovered from traps, suggesting that survivingites may have been newly moulted first instar nymphs

hat had yet to be fully affected by product application.hese observations further support that a repeat applica-ion of spinosad 7–14 days after initial treatment would beeneficial to maximise control efficacy.

Estimating the exact population of D. gallinae in a com-ercial housing system is an extremely difficult task, since

actors including the position of the trap and distributionf mite refugia will influence the exact number of miteshat take refuge in the trap. This may be one of the reasonshy population surveys of D. gallinae in commercial hous-

ng systems have shown extremely high variation in theumber of mites recovered per trap (e.g. Guy et al., 2004).evertheless, the methodology used in the current exper-

ment ensured that the protocol for population samplingas same for each cage/cage group, so that the effect of

reatment seen can be considered reliable.

. Conclusion

This study has confirmed both the in vitro and in vivooxicity of spinosad to D. gallinae and demonstrated thatffective control of D. gallinae can be achieved for a period oft least 28 days after spraying. Although statistically thereas no difference in efficacy between the two application

ates used in the in vivo study, the higher application rate ofpinosad consistently provided ∼20% greater reductions in. gallinae numbers. This suggests that an application ratef 3.88 g/L of spinosad sprayed to the point of run-off woulde most appropriate for commercial use, where maximum. gallinae mortality should be sought to minimise thehance of resistance developing. Reassuringly, there was novidence of product application affecting either hen body-eight or egg production, suggesting that spinosad could

e used whilst birds are in lay and thus give commercial eggroducers an additional facet to an integrated managementtrategy for D. gallinae.

cknowledgements

The authors would like to thank the staff of Roslinutrition Ltd., in particular Mr Doug Currie and Mr Brianobertson, for assistance with experimental planning, sitereparation and day-to-day running of the in vivo study.

eferences

nastas, P., Kirchchoff, M., Williamson, T., 1999. 1999 Green Chemistryawards: spinosad—a new natural product for insect control. GreenChem. 1, G88.

rkle, S., Guy, J.H., Blackett, S.M., Sparagano, O., 2004. Variation in the

population of Dermanyssus gallinae in a free range laying unit andeffectiveness of chemical control. Br. Poult. Sci. 45, S45–S46.

rkle, S., Guy, J.H., Sparagano, O., 2006. Immunological effects and pro-ductivity variation of red mite (Dermanyssus gallinae) on layinghens—implications for egg production and quality. World Poult. Sci. J.62, 249–257.

tology 173 (2010) 307–316 315

Beugnet, F., Chauve, C., Gauthey, M., Beert, L., 1997. Resistance ofthe red poultry mite to pyrethroids in France. Vet. Rec. 140,577–579.

Cetin, H., Cilek, J.E., Oz, E., Aydin, L., Deveci, O., Yanikoglu, A., 2009. Compar-ative efficacy of spinosad with conventional acaricides against hardand soft tick populations from Antalya. Turkey Vet. Parasitol. 163,101–104.

Chauve, C., 1998. The poultry red mite Dermanyssus gallinae (De Geer,1778): current situation and future prospects for control. Vet. Para-sitol. 79, 239–245.

Chirico, J., Eriksson, H., Fossum, O., Jansson, D., 2003. The poultry red mite,Dermanyssus gallinae, a potential vector of Erysipelothrix rhusiopathiaecausing erysipelas in hens. Med. Vet. Entomol. 17, 232–234.

Cosoroaba, I., 2001. Massive Dermanyssus gallinae invasion in battery-husbandry raised fowls. Revue Méd. Vét. 152, 89–96.

Davey, R.B., Miller, J.A., George, J.E., Snyder, D.E., 2005. Effect of repeatedspinosad treatments on cattle against Boophilus annulatus under SouthTexas field conditions. Southwestern Entomol. 30, 245–255.

Davey, R.B., George, J.E., Snyder, D.E., 2001. Efficacy of a single whole-body spray treatment of spinosad, against Boophilus microplus (Acari:Ixodidae) on cattle. Vet. Parasitol. 99, 41–52.

Fiddes, M.D., Le Gresley, S., Parsons, D.G., Epe, C., Coles, G.C., Stafford, K.A.,2005. Prevalence of the poultry red mite (Dermanyssus gallinae) inEngland. Vet. Rec. 157, 233–235.

George, D.R., Olatunji, G., Guy, J.H., Sparagano, O.A.E., 2010. Effect of plantessential oils as acaricides against the poultry red mite, Dermanys-sus gallinae, with special focus on exposure time. Vet. Parasitol. 169,222–225.

Guy, J.H., Khajavi, M., Hlalele, M.M., Sparagano, O., 2004. Red mite (Der-manyssus gallinae) prevalence in laying units in Northern England. Br.Poult. Sci. 45 (Supplement 1), S15 (Abstr.).

Holt, K.M., Opit, G.P., Nechols, J.R., Margolies, D.C., 2006. Testing for non-target effects of spinosad on two-spotted spider mites and theirpredator Phytoseiulus persimilis under greenhouse conditions. Exp.Appl. Acarol. 38, 141–149.

Keïta, A., Pagot, E., Pommier, P., Baduel, L., Heine, J., 2006. Efficacyof phoxim 50% E.C. (ByeMite) for treatment of Dermanyssus gal-linae in laying hens under field conditions. Revue Méd. Vét. 157,590–594.

Kilpinen, O., Steenberg, T., Jespersen, J.B., Birkett, M., Wadhams, L., Pickett,J., Cruz, M.D.S., 2002. Development of new control methods againstchicken mites. In: Good, M., Hall, M.J., Losson, B., O’Brien, D., Pithan,K., Sol, J. (Eds.), Mange and Myiasis of Livestock –Proceedings of theWorkshop Held at the École Nationale vétérinaire de Toulouse. France,3rd–6th October 2001, pp. 33–35.

Kim, S., Yi, J., Tak, J., Ahn, Y., 2004. Acaricidal activity of plant essential oilsagainst Dermanyssus gallinae (Acari: Dermanyssidae). Vet. Parasitol.120, 297–304.

Meyer-Kühling, B., Pfister, K., Müller-Lindloff, J., Heine, J., 2007a. Fieldefficacy of phoxim 50% (ByeMite®) against the poultry red mite Der-manyssus gallinae in battery cages stocked with laying hens. Vet.Parasitol. 147, 289–296.

Meyer-Kühling, B., Heine, J., Müller-Lindloff, J., Pfister, K., 2007b. Epidemi-ology of Dermanyssus gallinae and acaricidal efficacy of phoxim 50%in alternative housing systems during the laying period of hens. Par-asitol. Res. 101, S1–S12.

Meyling, N.V., Eilenberg, J., 2007. Ecology of the entomopathogenic fungiBeauveria bassiana and Metarhizium anisopliae in temperate agroe-cosytems: potential for conservation biological control. Biol. Control43, 145–155.

Nordenfors, H., Höglund, J., 2000. Long term dynamics of Dermanyssusgallinae in relation to mite control measures in aviary systems forlayers. Br. Poult. Sci. 41, 533–540.

Sparagano, O., Pavlicevic, A., Murano, T., Camarda, A., Sahibi, H., Kilpinen,O., Mul, M., van Emous, R., le Bouquin, S., Hoel, K., Cafiero, M.A.;1;,2009. Prevalence and key figures for the poultry red mite Dermanys-sus gallinae infections in poultry farm systems. In: Sparagano, O.A.E.(Ed.), Control of poultry mites (Dermanyssus). Exp. Appl. Acarol. 48,3–10.

Thompson, G.D., Dutton, R., Sparks, T.C., 2000. Spinosad: an examplefrom a natural products discovery programme. Pest Manage. Sci. 56,696–702.

Urquhart, G.M., Armour, J., Duncan, J.L., Dunn, A.M., Jennings, F.W., 1996.

Veterinary Parasitology, 2nd ed. Blackwell Scientific Publications,Oxford, UK.

Valiente Moro, C., De Luna, C.J., Tod, A., Guy, J.H., Sparagano, O.A.E.,Zenner, L., 2009, The poultry red mite (Dermanyssus gallinae): apotential vector of pathogenic agents. In: Sparagano, O.A.E. (Ed.),Control of poultry mites (Dermanyssus). Exp. Appl. Acarol. 48,93–104.

316 D.R. George et al. / Veterinary Parasi

van Emous, R., 2005. Wage war against the red mite! Poult. Int. 44, 26–33.Villanueva, R.T., Walgenbach, J., 2006. Acaricidal properties of spinosad

against Tetranychus urticae and Panonychus ulmi (Acari: Tetranychi-dae). J. Econ. Entomol. 99, 843–849.

tology 173 (2010) 307–316

Wojcik, A.R., Greygon-Franckiewicz, B., Zbikowska, E., Wasielewski, L.,2000. Invasion of Dermanyssus gallinae (De Geer, 1778) in poultryfarms in the Torun region. Wiad Parazytol. 46, 511–515 (in Polish withEnglish abstract).