Field Evaluation of Deltamethrin and Ivermectin Applications to Cattle on Culicoides Host-Alighting, Blood-Feeding, and Emergence

The impact of topical applications of deltamethrin and ivermectin to cattle on Culicoides spp. landing and blood-feeding was studied in this work using sticky traps mounted on Friesian heifers’ backs. There was no effect of the insecticides on total numbers of Culicoides trapped or the proportion engorged. Deltamethrin and ivermectin treatment did not prevent blood-feeding on these animals. Deltamethrin did result in significant Culicoides mortality as evidenced by the numbers of dead midges combed from heifers’ upper flanks. The proximity of engorged midges on traps to dead midges in the hair suggests that blood-feeding took place despite midges receiving an ultimately lethal dose of deltamethrin. Ivermectin application resulted in a smaller proportion of nulliparous than parous females caught. There was no significant effect of ivermectin on the numbers of Culicoides that emerged from dung samples (but p was small at 0.095 for the Obsoletus group Culicoides). In cases of suspect animal imports, pour-on or spray applications of deltamethrin could reduce the risk of onward transmission of bluetongue virus.


Introduction
Biting midges, Culicoides spp. (Diptera: Ceratopogonidae), are important vectors of viral pathogens of livestock. During 2006-2009, there was a serious outbreak of the midge-vectored virus, bluetongue virus (BTV) serotype 8, in north-western Europe. The virus causes a notifiable disease and the outbreak resulted in widespread disruption to the cattle and sheep livestock industries. Whilst vaccination was the ultimate solution to the bluetongue outbreak, at early stages of a disease outbreak vaccines may not be developed or available in sufficient quantities for disease management. At these crucial early stages, vector control is a valuable option to prevent spread of the disease [1]. The island of Ireland has remained free from bluetongue, and did so even during the 2006-2009 European outbreak. The main risk of incursion has been from imported livestock [2]. For example, in November 2018, four heifers and one bull were imported into a Northern Irish farm from France. Routine post-import testing found that one of the heifers tested PCR positive for BTV-8 [3]. The infected heifer was euthanized and the other animals quarantined. This was during the 'vector-free' period as defined by fewer than five parous Culicoides per light trap catch [4], but small numbers of midges were still flying. In cases such as this, insecticide application could reduce the likelihood of onward transmission of the pathogen.
Application of insecticides directly to animals is a common method of nuisance fly management. Two of the most commonly used insecticide groups approved for nuisance fly and ectoparasite control on midges 35 days after treatment, whilst Venail et al. [29] predicted maximum mortality after 4 days but that the lethal effect could be as brief as 10 days. For ivermectin, maximum blood concentration occurs 3-4 days after topical application [30,31]. However, in a study on horses the concentration of ivermectin in hair at the pour-on application site was well above that likely to cause Culicoides mortality (i.e., 350 ng per mL [22]) at greater than 10 µg per g for at least 40 days [31]. Therefore, the first trial in this study looked at trapping midges for a prolonged period (5 weeks) post application, whilst the following two trials concentrated on a two-week period post application when effects on midges were most likely. In the main, two hypotheses were tested in these trials. The first was that treatment with insecticidal compounds would prevent midge alighting (repellency). The second was that treatment would prevent blood-feeding. In addition, dung samples were taken to assess the effects of drug residues on Culicoides emergence from dung.

Material and Methods
Studies were conducted at the Agri-Food and Biosciences Institute's (AFBI) Hillsborough research farm, Co. Down, Northern Ireland (54.445290 • N, 6.065526 • W). The experiments were conducted May-July 2012 (Trial 1) and September 2014 (Trial 2 and 3).

Trial 1-Five Week Study on the Effects of Deltamethrin and Ivermectin on Culicoides' Landing and Emergence from Dung
The first trial used ivermectin (as Ivomec Classic Pour-On for Cattle, Boehringer Ingelheim Animal Health UK Ltd., Bracknell, UK) and deltamethrin (as Fly and Lice Spot On Insecticide, Zoetis UK Ltd., London, UK). Fifteen Friesian heifers (live weight~370 kg) were selected from the farm's herd at random and were allocated to three treatments: five were used as a control, five were treated with ivermectin, and five were treated with deltamethrin one day prior to the start of Culicoides monitoring. Animals were treated according to the manufacturers' instructions. In brief, for ivermectin 1 mL product per 10 kg bodyweight (500 µg ivermectin per kg bodyweight) was applied along the mid-line of the back in a narrow strip between the withers and tailhead. For the deltamethrin, a standard 10 mL of product irrespective of animal weight (0.1 g deltamethrin) was applied as a single dose on the mid-line of the back, at the shoulders.
To monitor Culicoides landing, each animal had two white 200 cm 2 sticky traps (Oecos Insect Monitoring, Kimpton, UK; Agralan Ltd, Swindon, UK) attached to its back using Velcro ® and Kamar ® adhesive glue (www.kamarinc.com) as per the method of Thompson et al. [25] (Figure 1). Traps were placed on the heifers 1 day after treatment. These traps were left in place for 24 h before being removed. This sampling was repeated at weekly intervals for a total of 5 weeks, equivalent to a period of 30 days after treatment. This first trial examined the persistence of any treatment effects over 5 weeks, whereas the two subsequent trials concentrated on multiple sampling within 15 days of application when insecticides were most bioavailable and effects most likely observed.
Culicoides caught on the sticky traps were left in situ. For Culicoides identification, specimens were not identified to species but were grouped primarily according to sub-genus and then sub-divided again by morphological characteristics, mainly wing patterning. This approach has been commonly used in the veterinary and applied studies of Culicoides [32,33] [34]. Additionally, female parity was assessed by abdominal pigmentation using the method described by Dyce [35].
Dung samples were taken from each of the animals, either by collecting from dung pats immediately after defecation or directly from the rectum. Dung samples (300 mL) were laid out in an adjacent paddock in a randomised block design with 1 m between treatments and left exposed to Culicoides oviposition for one week before being covered with a bucket emergence trap for 4 weeks to allow for emergence [36]. Following this four week period, the bucket traps were removed and Culicoides counted. A soil core (diameter 10 cm and depth 6 cm) was then taken from each dung sample and put into a breathable tissue-culture bag (450 × 200 × 120 mm, Sigma-Aldrich) containing a 100 cm 2 sticky trap and incubated for a further four weeks at 20 • C (light regime 16 h light, 8 h dark) to determine any midge emergence. Any Culicoides caught on the sticky traps were identified to group level and counted. The first dung samples were taken one day after treatment, with sampling repeated at weekly intervals for a total of 5 weeks, resulting in 75 dung samples in total (3 treatments × 5 replicates × 5 weeks); however, the initial samples were destroyed after cattle gained access to the paddock and trampled the emergence traps. to allow for emergence [36]. Following this four week period, the bucket traps were removed and Culicoides counted. A soil core (diameter 10 cm and depth 6 cm) was then taken from each dung sample and put into a breathable tissue-culture bag (450 × 200 × 120 mm, Sigma-Aldrich) containing a 100 cm 2 sticky trap and incubated for a further four weeks at 20 °C (light regime 16 h light, 8 h dark) to determine any midge emergence. Any Culicoides caught on the sticky traps were identified to group level and counted. The first dung samples were taken one day after treatment, with sampling repeated at weekly intervals for a total of 5 weeks, resulting in 75 dung samples in total (3 treatments × 5 replicates × 5 weeks); however, the initial samples were destroyed after cattle gained access to the paddock and trampled the emergence traps.

Trial 2-Culicoides' Landing and Blood-Feeding on Friesian Heifers Treated with Ivermectin Pour-On
Twenty Friesian heifers (live weight ~350 kg) were treated with two antihelmintics. Ten were treated with ivermectin (as Ivomec) as per trial 1, and the other 10 with fenbendazole. The latter treatment equates to the control as animal husbandry practices on the farm did not allow an untreated control at this time but required the use of an antihelmintic against gastrointestinal parasites. Fenbendazole, as Panacur® (MSD Animal Health, Walton, UK), was applied as a 10% oral suspension at 1 mL of the product per 13 kg bodyweight (7.5 mg fenbendazole per kg bodyweight). Five days after treatment, two white sticky traps (200 cm 2 ) were mounted to the heifers' backs for a 24 h period, a process which was repeated again the following day (Table 1). Total numbers of Culicoides and those blood-fed were counted immediately after collection.

Trial 2-Culicoides' Landing and Blood-Feeding on Friesian Heifers Treated with Ivermectin Pour-On
Twenty Friesian heifers (live weight~350 kg) were treated with two antihelmintics. Ten were treated with ivermectin (as Ivomec) as per trial 1, and the other 10 with fenbendazole. The latter treatment equates to the control as animal husbandry practices on the farm did not allow an untreated control at this time but required the use of an antihelmintic against gastrointestinal parasites. Fenbendazole, as Panacur ® (MSD Animal Health, Walton, UK), was applied as a 10% oral suspension at 1 mL of the product per 13 kg bodyweight (7.5 mg fenbendazole per kg bodyweight). Five days after treatment, two white sticky traps (200 cm 2 ) were mounted to the heifers' backs for a 24 h period, a process which was repeated again the following day (Table 1). Total numbers of Culicoides and those blood-fed were counted immediately after collection. Table 1. Timing of treatments (days after application of ivermectin or deltamethrin) and sampling of heifers for Culicoides using sticky traps and combing in two consecutive trials in a Northern Irish pasture. The ivermectin-treated animals from trial 2 were also sampled in trial 3.

Trial 3-Culicoides' Landing, Blood-Feeding and Mortality on Friesian Heifers Treated with Deltamethrin and Ivermectin Pour-Ons
Following on directly from trial 2, one week after the initial treatments of ivermectin and fenbendazole, five of the fenbendazole (control) heifers were treated with deltamethrin (as 'Fly and Lice Spot On Insecticide') to give an experiment comprising three treatments, i.e., deltamethrin (plus fenbendazole), ivermectin and fenbendazole alone, with five replicate animals ( Figure 2). Four days after deltamethrin application, sticky traps were mounted on the backs of heifers for four consecutive 24 h periods (Table 1). In addition to sticky trap sampling, to collect dead Culicoides in the heifers' hair, the upper half of the flanks of each heifer were combed from the withers to hips with a plastic head lice comb (teeth 14 mm length,~0.3 mm apart) (Superdrug, London, UK) on three consecutive days (Table 1). Combing was done by the same operator each time, taking care to ensure that the areas combed were comparable between animals. The relationship between the treatments in trial 2 and trial 3. Five heifers selected at random from the trial 2 ivermectin treatment were kept on for trial 3. The 10 heifers in the trial 2 control treatment were split into two groups of five for trial 3. One group was left as a control, whilst the other group was treated with deltamethrin.

Analyses
Culicoides count data were subjected to generalized linear mixed models (GLMMs) fitted with Figure 2. The relationship between the treatments in trial 2 and trial 3. Five heifers selected at random from the trial 2 ivermectin treatment were kept on for trial 3. The 10 heifers in the trial 2 control treatment were split into two groups of five for trial 3. One group was left as a control, whilst the other group was treated with deltamethrin.

Analyses
Culicoides count data were subjected to generalized linear mixed models (GLMMs) fitted with Poisson distributions and logarithmic link functions. Where proportion data were analysed, binomial distributions coupled with logit link functions were used. In the GLMMs, individual animals were modelled as random effects. The significance of the fixed effects in the models was assessed by comparing Wald statistics for each term against an appropriate F-distribution. All analyses were conducted using the statistical package GenStat v16.2 (VSN International Ltd, UK; www.vsni.co.uk).

Trial 1-Five Week Study on the Effects of Ivermectin and Deltamethrin on Culicoides' Landing and Emergence from Dung
The total number of Culicoides caught on sticky traps during this study was 1,889. All were female and belonged to the Obsoletus (67%) or Pulicaris (33%) groups. The numbers of Culicoides caught declined as the experiment progressed ( Figure 3). However, there was no effect of deltamethrin or ivermectin treatment on the number of Culicoides caught on sticky traps (overall Culicoides per heifer, back-transformed (b-t) from GLMM predictions with 95% confidence intervals: Obsoletus group − control  Two-hundred and eighty-three Culicoides were reared from dung collected from the experimental heifers, with these being 106 female and 177 male and predominantly Obsoletus group (95%) and Pulicaris group (5%), with a single midge from the Impunctatus group (<1%). The bucket emergence traps collected 116 midges (41%) and the incubated soil cores 167 (59%). There was no effect of ivermectin or deltamethrin on Culicoides emergence at the 5% significance level (Wald statistic = 2.75, d.f.

Trial 3-Culicoides' Landing, Blood-Feeding and Mortality on Friesian Heifers Treated with Deltamethrin and Ivermectin Pour-Ons
A total of 19,720 Culicoides (a maximum of 641 on a single trap) were caught on sticky traps on the 15 heifers over the four consecutive 24 h periods. Again, there was no significant effect of ivermectin or deltamethrin on the number of Culicoides caught on animal-mounted sticky traps (Wald statistic = 0.95, d.f. = 2, 13.5, p = 0.633; Figure 5) or the proportion of midges that had blood-fed (b-t means, proportion blood-fed, control = 0.11 (0.08-0.14), ivermectin = 0.14 (0.11-0.18), deltamethrin = 0.12 (0.10-0.16); Wald statistic = 2.63, d.f. = 2, 11.2, p = 0.307). A total of 19,720 Culicoides (a maximum of 641 on a single trap) were caught on sticky traps on the 15 heifers over the four consecutive 24 h periods. Again, there was no significant effect of ivermectin or deltamethrin on the number of Culicoides caught on animal-mounted sticky traps (Wald statistic = 0.95, d.f. = 2, 13.5, p = 0.633; Figure 5) or the proportion of midges that had blood-fed (b-t means, proportion blood-fed, control = 0.11 (0.08-0.14), ivermectin = 0.14 (0.11-0.18), deltamethrin = 0.12 (0.10-0.16); Wald statistic = 2.63, d.f. = 2, 11.2, p = 0.307). A total of 737 dead Culicoides were combed from the hair of the 15 experimental heifers from samples taken on two consecutive days. There were significantly more Culicoides removed from the hair of deltamethrin-treated heifers than those treated with ivermectin or the control heifers (Wald statistic = 19.71, d.f. = 2, 13.1, p = 0.002; Figure 6). Due to the dehydrated condition of these midges it was not possible to determine visually if they were blood-fed. A total of 737 dead Culicoides were combed from the hair of the 15 experimental heifers from samples taken on two consecutive days. There were significantly more Culicoides removed from the hair of deltamethrin-treated heifers than those treated with ivermectin or the control heifers (Wald statistic = 19.71, d.f. = 2, 13.1, p = 0.002; Figure 6). Due to the dehydrated condition of these midges it was not possible to determine visually if they were blood-fed.

Discussion
To use insecticides to protect livestock from Culicoides-vectored pathogens requires that the relationships between midges alighting on the host, blood-feeding, insecticide-repellence and insecticide-toxicity are understood. In these three trials, there was no effect of ivermectin or deltamethrin on the numbers of Culicoides caught on animal-mounted sticky traps, indicating that these treatments did not deter landing. This is perhaps to be expected as the main actions of deltamethrin and ivermectin are toxicity, although there is also some evidence of contact irritancy and repellence to deltamethrin [18,37,38] and ivermectin has reduced the responses of Culicoides imicola Kieffer, 1913 to host cues in an olfactometer [39]. Crucially, there was also no difference detected in the proportion of Culicoides blood-fed between treatments. The midges caught in this study were fresh blood-fed, which implied that they had fed on the heifers on which they were

Discussion
To use insecticides to protect livestock from Culicoides-vectored pathogens requires that the relationships between midges alighting on the host, blood-feeding, insecticide-repellence and insecticide-toxicity are understood. In these three trials, there was no effect of ivermectin or deltamethrin on the numbers of Culicoides caught on animal-mounted sticky traps, indicating that these treatments did not deter landing. This is perhaps to be expected as the main actions of deltamethrin and ivermectin are toxicity, although there is also some evidence of contact irritancy and repellence to Viruses 2019, 11, 731 9 of 13 deltamethrin [18,37,38] and ivermectin has reduced the responses of Culicoides imicola Kieffer, 1913 to host cues in an olfactometer [39]. Crucially, there was also no difference detected in the proportion of Culicoides blood-fed between treatments. The midges caught in this study were fresh blood-fed, which implied that they had fed on the heifers on which they were trapped. After engorgement, haematophagous insects leave the host to digest their blood meal. Therefore, this result indicates that the deltamethrin and ivermectin treatments did not prevent blood-feeding.
Maclachlan and Mayo [5] suggest that insecticides which allow Culicoides to blood-feed are not effective in bluetongue control as feeding midges can transmit the virus before dying. Robin et al. [19] found that a topical application of deltamethrin did not prevent blood-feeding on horses but did consider that if such treatments killed midges they could have a role in reducing onward transmission of disease from viraemic horses or suppressing the immediate Culicoides population. The combing technique in the present study showed that deltamethrin was killing midges that alighted on the heifers' upper flanks. Deltamethrin is lipophilic and is disseminated by the natural oil secretions of the coat, with the midges exposed to the insecticide as they crawl through the hair. Deltamethrin is highly toxic to European C. obsoletus, with an LD50 of 1.33 × 10 −4 % [40] but despite this, blood-feeding was not prevented. Other studies have found that the insecticide concentration declines from the mid-line of the back to the belly, legs or face [6,15,30,41]. However, in this study, since engorged Culicoides were trapped on the middle of the back of the heifers close to the deltamethrin deposition point and surrounded by dead midges in the hair, it is reasonable to assume that these midges were able to blood-feed despite receiving a subsequently lethal dose of insecticide. This differs from the conclusions of earlier studies exposing hair from treated animals to Culicoides. These considered that the rapid knockdown of Culicoides by deltamethrin would prevent blood-feeding [42,43].
Ivermectin has a different mode of action to deltamethrin. It is highly lipophilic but also systemic, being absorbed through the skin to the subcutaneous fat reserves and bloodstream. It may also bind to the hair and be retained on the skin [31]. There was no evidence of Culicoides mortality in this study but this cannot be discounted as ivermectin does not have the same knockdown effects as deltamethrin and midges may fly from the animal before dying. Furthermore, sub-lethal effects may have a role in suppressing disease transmission through reduced lifespan, fecundity, dispersal or altered vector-pathogen interactions. A good example of this is that ivermectin reduced the bluetongue (BTV-17) infection rate of C. sonorensis by 40% with a 29% reduction in dissemination of the virus from the midges' bodies to the heads [23]. The only significant effect of ivermectin on Culicoides in this study was a reduction in the proportion of nulliparous midges caught. The reasons for this are unknown. Different physiological states can influence how vectors respond to host cues. It is possible that nulliparous females are more sensitive to stimuli than older parous females [44] and are deterred by subtle changes in the host olfactory profile brought about by ivermectin treatment [39].
Ivermectin is excreted in the animals' dung and can have negative effects on dung-dwelling fauna [45][46][47], including nuisance and biting fly larvae [48,49]. In this study, the numbers of Culicoides that emerged from the dung samples was small. There was a tentative suggestion of a negative effect of ivermectin on emergence of the Obsoletus group, the main vector risk group, but this was not significant at the 5% level (p = 0.095). Nevertheless, ivermectin excreted in the dung is not considered a viable widespread strategy to reduce Culicoides population levels due to the adverse effect on beneficial non-target invertebrates and also the availability of other larval breeding sites for Culicoides. Deltamethrin may also have residual insecticidal activity in dung [50,51].
The technique of using sticky traps to provide a standardised method of assessing Culicoides landing on experimental animals generally worked well. The numbers of Culicoides caught on traps were variable but in the second and third trials in September, they were substantial. In the third trial, a mean control of 172 (maximum of 1260) Culicoides were caught per 200 cm 2 trap over a 24 h period. It is unknown to what extent these traps were attracting Culicoides to land on them compared to the surrounding animal's skin. However, if such figures are extrapolated to the surface area of the heifers, and allowing for preferential feeding on different areas of the body, then animals in this location face a biting intensity of tens of thousands of Culicoides per night. One disadvantage of sticky traps is that specimens are distorted and difficult to identify to species level. Light trap catches at this location were mostly Obsoletus group [52], a subsample of which were subsequently identified to species level with C. obsoletus s.l./scoticus, Culicoides lupicaris and Culicoides dewulfi predominant (Bruno Mathieu and Thomas Balenghien, pers. comm.) Although UV light trapping may overestimate the biting rates of C. obsoletus, they nonetheless give a reasonable indication of the species involved [53].

Conclusions
The main result from this study is that neither deltamethrin nor ivermectin prevented blood-feeding by Culicoides midges on cattle in a field situation. The implications of this are that these treatments do not have a protective effect for individual animals but could prevent onward transmission of pathogens via post-feeding mortality of the vector. This would certainly seem to be the case for deltamethrin, where large numbers of dead Culicoides were collected on the upper flanks of the animals. For ivermectin, it was not possible to assess the mortality of affected Culicoides as they most likely died off the animal. Therefore, at the present time, in cases of suspect imported livestock, which represent the most significant risk of bluetongue incursion into Ireland, application of deltamethrin as a pour-on or spray treatment applied to the quarantine animals would be advocated to reduce the risk of onward transmission. This should be combined with spraying of the housing using a residual insecticide and where possible screening of the entrances. Any native livestock on the farm should be situated as far as possible from the quarantine housing, although it should be considered that Culicoides can disperse up to 5 km within a few days [54].
Although the sticky trap technique worked well, the position of the traps was restricted to the mid-back where the heifers could not lick or swat the traps with their tails. The coverage provided by pour-on treatments diminishes with distance from the application site, so in a future study it would be valuable to assess Culicoides mortality in distal areas with maximal midge feeding, such as the underbelly [26], possibly using a sticky substance applied directly to the hair rather than sticky plastic plates.