First published in Animal Production Science 2012, 52 (7) 659-664 ( CSIRO Publishing )
In response to suggestions that the incidence of louse infestations in New South Wales has increased markedly, a survey of 173 producers was conducted in the Tablelands Livestock Health and Pest Authority region using visual detection and a questionnaire to document retrospective lice history. An estimated apparent prevalence of 16.5% is a moderate increase from the 10% reported in 2004. On a subset of the surveyed sheep flocks sheep, lice-specific immunoassay conducted by the New South Wales Department of Primary Industries were used to detect low levels of infestation that were not identified by visual detection. This provided a true prevalence estimate of 30%. These results will be used to promote improved control and preventative strategies.
An Australia-wide sheep lice (Bovicola ovis) survey in 2004 estimated prevalence of 13% in the high rainfall and cereal sheep zones, 21% in the pastoral zone and 20% in prime lamb flocks (Walkden-Brown et al. 2006). The cost of lice to the Australian wool industry is estimated to be $123 million (Sackett et al. 2006) due to fleece derangement caused by pruritis and the considerable costs associated with eradication and control (Wilkinson et al. 1982; James and Moon 1998; Sackett et al. 2006). This ranks lice as the third most economically important parasite of sheep following gastrointestinal helminths and sheep blowflies (James et al. 2002a).
There has been widespread speculation that the incidence of louse infestations in New South Wales flocks may be as high as 70% (B. R. Watt, pers. comm.) with Livestock Health and Pest Authority (LHPA) district veterinarians and NSW Primary Industries livestock officers reportedly expressing concerns about increasing infestations reaching intolerable levels in the high rainfall areas of NSW (Littlejohn and Schroder 2009). This perceived increased incidence may be due to emerging resistance to insect growth regulator insecticides (James et al. 2008; Levot and Sales 2008), poor application of lice control chemicals (Levot 2008), and/or inadequate on-farm bio-security (Joshua et al. 2010). However, in the absence of recent survey data this perceived increased prevalence may be exaggerated.
There have been extensive changes in the approach to lice control within each state in Australia (James 2010). In NSW, before the 1980s a regulatory approach legislated quarantine and stock movement restrictions for properties with known infestations and annual treatment of all flocks. In recent years, regulation was replaced with an advisory program with greater emphasis on education programs and a reduction in the use of chemical control and eradication methods (James and Riley 2004). Current advice is not to treat sheep unless lice are present (Horton et al. 2009).
Perhaps as a legacy of this regulatory approach, advisers have found it increasingly difficult to engage producers and deliver crucial lice control information (J. Eppleston, pers. comm.). To compound this problem, there are difficulties detecting sheep lice due to the low sensitivity of visual fleece partings (Pearse and Gardner 1994; James and Moon 1998). Between June 2009 and June 2011 an enzyme-linked immuno-sorbent assay (ELISA), developed collaboratively by CSIRO Livestock Industries, Australian Wool Innovation and the NSW Department of Primary Industries provided a highly sensitive supplement to visual detection (Sales and Young 2009), but limited utilisation by producers led to withdrawal of the test.
Monitoring changes in the prevalence of sheep lice would enable current control strategies to be reviewed and if necessary promote active engagement of sheep producers to re-evaluate their strategies to lice control. This paper reports the prevalence and distribution of louse infestations in sheep flocks in the central and southern Tablelands region of NSW measured using visual detection as well as the ELISA lice detection test.
The study was conducted within the Tablelands Livestock, Health and Pest Authority (TLHPA) area of NSW (Fig. 1). Major towns include Lithgow, Bathurst, Blayney, Oberon, Orange, Yass, and Goulburn. The region is categorised by annual rainfall of ~650 mm (Bureau of Meteorology 2010) and an elevated topography of 500-1000 m above sea level (NSW Agriculture 2003). It is predominately utilised by sheep, wool and cattle grazing enterprises. At the time of the study the TLHPA area ran ~2.55 million sheep (B. R. Watt, pers. comm.).
A random number generator was used to select 450 producers from the TLHPA database of 2712 enterprises running more than 50 sheep. These producers were sent a letter describing the survey and reasons for its conduct followed by telephone contact to determine their willingness to participate. One-hundred and seventy-three producers agreed to participate and indicated the date they intended to commence shearing. The study was granted University of Sydney Human Ethics Committee approval: Protocol No. 12802 and NSW I&I Orange Animal Ethics Committee approval No. ORA 10/005.
Sheep lice detection
Inspections of sheep flocks for the presence of lice were conducted between September 2009 and October 2010 by one of six TLHPA animal health staff members trained in visual detection and collection of samples to be analysed using the lice detection ELISA. Initially, a visual assessment was employed on the most 'at-risk' flock at each of the 173 properties included in the study. At-risk flocks are as defined by James et al. (1998) as: the youngest animals; where straying may have occurred from neighbouring properties; and where there were known previous louse infestations. The most 'at-risk' flock on each property was determined in consultation with the property manager. Where no lice were detected by visual inspection during the period July to October 2010 a lice detection ELISA was utilised to estimate the sensitivity of visual observations for 24 flocks where the flock size was in excess of 200 sheep.
Visual detection. This inspection followed the protocol outlined by Joshua et al. (2010). Briefly, the flock of interest was first observed from a distance to detect sheep showing signs of pruritic behaviour such as rubbing, biting or scratching. The flock was then run slowly between two observers at least twice to look for sheep with signs of fleece derangement. Ten sheep showing signs of fleece derangement were inspected. If unable to detect sheep with fleece derangement, 10 sheep from the flock were selected and inspected. The inspection for lice on individual sheep was performed by parting the fleece ~10 cm to expose the skin at 10 sites (5 partings on each side) along the upper midline between the neck and flank on the sheep. The sheep were positioned so that the opened fleece partings were exposed to the best possible light, preferably direct sunlight. If direct sunlight was unavailable artificial light was used. If one louse was found on one sheep the whole flock was considered to be infested. The numbers of lice found at each 10-cm parting were used to estimate the severity of infestation. This was derived by averaging the total amount of lice found in each fleece parting and the degree of infestation categorised as nil (0 louse per parting), low (<2 louse per parting), medium (2-5 louse per parting) or high (>5 louse per parting) (NSW Agriculture 2001).
ELISA samples. For 24 flocks, where lice were not visually detected, a sample collection kit was used following the protocol outlined by Sales and Young (2009). Briefly, all combs and cutters were ensured to be scrupulously clean before shearing ~200-250 sheep from the flock of interest. Washes were collected from combs and cutters using a kit containing a bottle with detergent, a dish and a scrubbing brush. The bottle was filled with potable water, shaken and the diluted detergent poured into the dish. The combs and cutters used to shear the sheep were soaked in the solution and scrubbed with the brush provided. The resulting washings were then poured back into the bottle which was then sealed. Each sample was posted to the Elizabeth Macarthur Agricultural Institute where, an ELISA was performed to detect a unique louse protein that indicated the presence of lice.
Questionnaire and spatial distribution
A questionnaire was completed with the owner or manager of the flock at the time of inspection. This included eight questions relating to flock management, enterprise structure and lice history for the period 2004-09. Specific questions relating to lice prevalence included: is the flock currently infested; were lice present during the proceeding 5-year period; if lice were detected, what treatment method and product were used? The risk of lateral spread was also examined with questions relating to sheep introductions, status of neighbouring flocks and completeness at each muster.
Map coordinates for the centre of each property surveyed were used to establish the spatial distribution of infested flocks across the TLHPA area.
Data management and analyses
The 'apparent prevalence' was calculated as the proportion of surveyed flocks that were classified as infested by visual detection or the results of the ELISA. The 'true prevalence' (Dohoo et al. 2009), the proportion of the population with lice within the central and southern tablelands region of NSW, was estimated using an on-line epidemiology prevalence calculator (AusVet Animal Health Services 2009). The input assumptions for the visual detection were: test sensitivity = 0.6; test specificity = 0.999; confidence level = 0.95 (James et al. 2002b). A trendline was added using MS Excel, including the equation and R-squared value, to demonstrate the change in lice prevalence over time.
Visual inspection of 173 flocks suggested that 28 (16.2%) were lousy. These comprised 16 farms with low infestation, eight farms with medium infestation and four highly infested farms. ELISA results for 4 of the 24 flocks assessed visually as louse-free, tested positive for lice. This increased the proportion of farms detected to have a louse infestation to 18.5% because four farms from this subsample tested positive.
The online Ausvet epidemiology prevalence calculator, with a determined sensitivity and specificity of 60 and 99%, respectively, estimated the true prevalence using visual detection alone to be 27% (0.18, 0.36 95% CI). When considering the proportion of False Negatives (4/24 : 16.7%) identified by ELISA results, the extrapolated apparent prevalence is estimated to be 52/173 (30%) if the same proportion occurred in the other 121 flocks visually assessed to be negative.
The spatial distribution of properties inspected for louse infestations within the TLHPA district of New South Wales is presented in Fig. 2. The distribution of louse infestations within the TLHPA area provides no clear pattern, other than that infested flocks have been detected throughout.
Retrospective information on apparent lice prevalence from 2004 to 2009 was obtained for the 173 properties (Fig. 3, F = 63.04, P = 0.02) suggesting an increasing trend in sheep lice prevalence in the central and southern Tablelands region.
The estimated true lice prevalence of 30% for the central and southern Tablelands region of NSW, although not as high as anecdotally suggested (B. R. Watt, pers. comm.), is in line with an estimated lice prevalence of 28 and 25% reported for the central and southern Tablelands, respectively, in 1981 by Roth and Plant (1992). This is considerably higher than the 10% indicated in the 2004 Walkden-Brown et al. (2006) study and the 11 and 12% reported for the same regions in 1991 (Roth and Plant 1992). However, different methodologies with different levels of sensitivity make comparisons difficult. Roth and Plant (1992) inspected fleeces in the woolstore whereas Walkden-Brown et al. (2006) conducted a nationwide postal survey. In the present study, the apparent prevalence of 16.5% is similar to the 'apparent' prevalence figures reported by Walkden-Brown et al. (2006), which are based on the owner's opinion as to whether the flock was louse infested. It could be argued that the present survey may have an increased sensitivity to the Walkden- Brown et al. (2006) postal survey as it used experienced independent observers and the sheep were almost always in full wool allowing nearly 12 months for lice to become detectable. Results from the present study suggest lice prevalence may have increased moderately over the past 6 years in the Tablelands region of NSW and is supported by the trend displayed in Fig. 3. James (2011) suggests two recent upsurges in louse prevalence have been associated with emerging resistance to previously highly successful (backline) products that commanded a dominant market share. The first upsurge commencing in the late 1980s coincided with the development and spread of resistance to the synthetic pyrethroid group of chemicals (James and Levot 2005). The most recent upsurge is associated with resistance to the insect growth regulators diflubenzuron and trifumuron (James et al. 2008; Levot and Sales 2008). However, poor product application, the withdrawal of diazinon due to concerns about human health risks from exposure (Department of Health and Aging 2011), recent widespread drought contributing to increased louse infestations in animals with poor nutrition or under stress (James 2010) and low wool prices discouraging sheep producers from spending time and money on lice control and management (Morcombe et al. 1994) including fencing have all contributed to an increased prevalence of lice.
It is possible the visual lice detection method employed in this study may provide an underestimate of true prevalence. Depending on factors such as the level of infestation and inspection regime, the sensitivity of detecting lice using live sheep inspections may be low (60%) (James et al. 2002b), even when high numbers of sheep per flock and fleece partings per sheep are inspected (Pearse and Gardner 1994; James and Moon 1998). Lice can be difficult to detect, as it has been estimated ~2500 lice per sheep must be present to observe an average of one louse per 10-cm wool parting (James and Moon 1999). In the early stages of an infestation only a few sheep in the flock are likely to have lice, with numbers of lice on each sheep also likely to be low due to a slow intrinsic rate of increase and spread (Murray and Gordon 1969). The participation of several TLPHA staff members with similar prior experience was not expected to underestimate prevalence beyond that expected from the inherent limitations in the method used.
It is important to acknowledge a voluntary survey should not be regarded as 'random' as the motivation for producers' participation or not is very likely driven by their perceptions of the survey outcomes. Given the propensity of neighbours to blame each other for louse infestation on their farm, it would be interesting to know if the incidence of lice occurred in clusters of neighbouring properties. Unfortunately, insufficient data was obtained in this study to tease out this information.
A supplement to visual lice detection was the sheep lice detection ELISA. At the time of the survey this test cost $133.41 per sample (Sales and Young 2009) and was only used on a subsample of flocks when lice were not visually detected. Although there was no concern with the sensitivity of the test, the cost, sampling technique and delay in receiving results may have discouraged producer uptake and led to the withdrawal of the test. Unless all sheep from the flock of interest were sampled the efficiency of the test could be reduced if only a few sheep in the flock were infested and these were not captured in the sample (James et al. 2002b). Moreover, a delay in receiving results is problematic if producers want to treat off-shears (within 24 h after shearing). The only other option if a positive result is confirmed is to treat in short wool but this requires a second muster. Future research aimed at the development of an on-farm test which would provide same day diagnosis would be advantageous.
The lice detection ELISA was able to detect low levels of infestation that were not apparent by visual detection. This is biologically and economically important, especially in terms of resistance and pest management (Levot 1995; Horton et al. 2009). Producers should not always rely on negative results from visual inspections to make a decision regarding lice control. In this study, 4 of the 24 flocks visually assessed to be negative and selected for objective testing using the ELISA tested positive. A lice detection test, such as that previously offered by the NSW Department of Primary Industries, would be a very useful tool in determining the presence of sub-clinical infestations.
It is uncertain whether the replacement of regulation with an advisory program has contributed to an increase in lice prevalence in the central and southern Tablelands region of NSW. Although a greater emphasis on education and a reduction in the use of chemical control has the potential for improving the effectiveness of control programs through extending the lifespan of the chemicals used (James and Riley 2004), this will only be effective if sheep producers are engaged in the process. It is likely that the extended drought, low wool prices and the high cost of lice control have all contributed to the disengagement of producers from lice management programs. This disengagement is likely to have exacerbated the situation further and contributed to an increase in lice prevalence. Furthermore, the increase in land subdivision and numbers of small 'hobby' flocks (Behrendt and Eppleston 2011) may make regional lice control more difficult.
Questionnaires have proven to be beneficial in estimating the prevalence of lice at a regional level (Roth and Plant 1992; Walkden-Brown et al. 2006). However, the prevalence estimated from the questionnaire is dependent on producers' retrospective knowledge of the presence of lice in their flocks. Due to the difficulties detecting lice in the early stages of an infestation this method may be an underestimate or overestimate of the true prevalence.
Overall, these results will provide useful information to the sheep and wool industry on the prevalence of lice in the Tablelands region of NSW. With the knowledge from this survey we hope that the TLHPA will reinvigorate the regional advisory program and will engage producers in a regional control program.
Comparison of the results of a survey, that included visual and immuno-diagnostic inspection of sheep flocks from the central and southern Tablelands region of NSW, with published historical estimates of lice prevalence, suggest that lice infestations may have increased moderately. Factors including resistance to the insect growth regulators, removal of diazinon, extended drought, low wool prices and the high perceived cost of control may have contributed to this. It is hoped that educating producers about the extent of the problem as well as the advantages of adoption of technologies, such as the lice detection ELISA, to confirm the absence or presence of lice, will facilitate improved regional control.
This project was funded by the Australian Wool Education Trust. The authors acknowledge TLHPA animal health staff Scott Craig, Chris Harris, Scott Schlunke, Dennis Ferson, and Neville Collins for inspecting flocks; Gary Levot, Narelle Sales and Paul Young from the Elizabeth Macarthur Agricultural Institute for assistance with training and analysis; and Jenny- Ann Toribio and Simone Firestone from the University of Sydney for their epidemiological and spatial expertise. Finally, the sheep producers involved in this study are acknowledged for their willingness to participate.