Event Abstract

Back to Event

Assessment of spatial, environmental, and management effects on lameness prevalence in UK sheep flocks

Jessica D. Witt1*
  • 1 University of Warwick, United Kingdom

1. Background and Objectives Footrot (FR) in sheep is an infectious disease documented for ~150 years (Youatt, 1869). FR is caused by Dichelobacter nodosus, a gram-negative aerotolerant anaerobe (Beveridge & Gregory, 1941). Other bacteria have been observed colonizing after infection, including Fusobacterium necrophorum and various Treponeme species, but their role is still under investigation (Beveridge & Gregory, 1941; Frosth, Konig, Nyman, Pringle, & Aspan, 2015; Roberts & Egerton, 1969; Witcomb et al., 2014). In the UK, FR is endemic and is the primary cause of lameness in most sheep flocks, accounting for ~70% of all reported lameness cases (Winter, Kaler, Ferguson, KilBride, & Green, 2015). Beveridge and Gregory (1941) attributed outbreaks of FR to lush pasture growth and postulated that an injury to the interdigital skin, potentially from the pasture, was necessary to establish the condition. Later studies support this, as Graham and Egerton (1968) observed little to no transmission during the dry season with no pasture growth and Woolaston (1993) reported that ewes that lambed in well-drained fields with long, mature grass had a significantly higher risk of developing FR. Within the UK, occurrence of FR was positively associated with the 2-week rolling total rainfall and negatively associated with the 2-week rolling mean minimum temperature (Smith et al., 2014). Occurrence of FR was highest in April when the flock was housed for lambing; a finding supported by a clinical trial of 44 English flocks (Witt & Green, 2018). A UK laboratory study also identified predominantly clay soil with consistent moisture levels at 5°C was ideal for bacterial survival, producing metabolically active D. nodosus for over 30 days (Muzafar et al., 2016). Most evidence on environmental factors affecting the spread of FR has come from countries with different climates and landscapes to the UK, so are not directly applicable. Laboratory evidence cannot be assumed to reflect real-world conditions. Spatial variations that may contribute to the spread of FR in the UK have not yet been evaluated. This study aims to evaluate spatial variation in annual lameness prevalence and assess the impact of static environmental factors and lameness management practices on lameness prevalence. 2. Methods Based on data from 2013 and 2014 submitted by farmers via questionnaires, 802 farms were selected that answered 64 questions of interest over both years. Postcodes for each farm were used to geocode the farm coordinates. Accuracy of the geocoding was tested using postcodes of 46 farms where the exact location was known. Of these 46 farms, 45 of the geocoded locations were within 2km or less of the actual farm site, giving an accuracy rate of 97.8%, which we felt was acceptable. Using ArcGIS 10.5.1 (2017), a 2km radius buffer zone was drawn around each geocoded location. These buffered zones became polygons within a single shapefile and were used for collection and calculation of farm area variables, including the soil types present, percent representation of soil components, area represented by each of 5 agricultural land classifications (ALC), and the maximum, minimum, and mean altitude. Soil type, soil components within each type, and area represented by each type within each polygon were provided by Cranfield University. Altitude measurements were provided by ESRI, and ALC data was provided by DEFRA via ESRI. 2.1 Spatial autocorrelation assessment using Local Moran’s I (LMI) statistic Distribution of sheep flocks within the UK is not homogeneous. The distribution of the farms selected for this analysis closely followed that of the national distribution. Because of this, and the continuous outcome variable of lameness prevalence, a Local Moran’s I (LMI) statistic (Anselin, 1995) was calculated. The LMI calculation evaluates a point’s similarity to its adjacent neighbours as well as the similarity of the adjacent neighbours to the whole study area. Four significant outcomes are possible – clusters of high prevalence (HH), clusters of low prevalence (LL), low outliers (LH) and high outliers (HL). The spatially weighted matrix was generated to include a set number of nearest neighbours based on their Euclidean distance from each point. For individual years, the nearest 8 neighbours were used. As the full data set contained 2 data points for each location, the spatially weighted matrix was based on the 17 nearest neighbours to ensure that data from the same 8 farm locations from the individual year analyses were used. To account for the spatial clustering inherent in the repeated measures data, a false discovery rate correction was applied to the analysis of the full data set. 2.2 Multi-level multivariable modelling of management practices Due to the skewed distribution of the lameness count data and the clustering of results by farm and year, a two-level over-dispersed Poisson model was constructed in MLwiN 2.35 (2014). Based on univariable results, the model was built in a forward stepwise fashion. Relative risks (RR) were calculated from the model coefficients and standard errors. Residuals of the final model were assessed for spatial autocorrelation using the same method as the full data set described above. 3. Results 3.1 Spatial autocorrelation results The 2013 LMI identified 69 low prevalence clusters, 38 low prevalence outlier farms, 7 high prevalence clusters, and 6 high prevalence outliers (Figure 1). The 2014 LMI identified 40 low prevalence clusters, 23 low prevalence outlier farms, 20 high prevalence clusters, and 18 high prevalence outliers (Figure 2). The full data set LMI identified 46 low prevalence clusters, 7 low prevalence outlier farms, 7 high prevalence clusters, and 9 high prevalence outliers (Figure 3). 3.2 Multi-level, multivariable modelling results Significant variables from the over-dispersed model are detailed in Table 1. Variables associated with a reduced RR of lameness included increases in flock size, increased land area that is ALC2 or ALC4, increased soil clay content, farms that don’t or are unsure if they have contagious ovine digital dermatitis (CODD), not separating ewes with footrot, avoiding foot trimming, and not putting ewes through a footbath. Variables associated with an increased RR of lameness included increased soil silt content, lameness prevalence in lambs, being unable to recognize lame sheep at score 1, incorrectly identifying FR lesion images, never using injectable antibiotics to treat ewes with footrot, not putting lambs through a footbath, not routinely checking feet of sheep before purchasing, not isolating animals returning to the flock, and not vaccinating ewes against FR (Footvax ™). The LMI analysis of the model residuals identified 11 low prevalence clusters, 4 low prevalence outliers, 2 high prevalence clusters and 2 high prevalence outliers (Figure 4). 4. Conclusions At farm level, there were fewer significant LMI results in 2014 compared to 2013, but there was a shift to a greater number of significant high prevalence results and fewer low prevalence results. The variability in average lameness prevalence was also decreased. Overall, there were more significantly high prevalence areas in 2014, but they were less extreme. Results of the full data set LMI have a similar distribution to the individual year results, though with far fewer significant low prevalence outliers than observed in 2013 or 2014 (7 compared to 38 and 23 respectively). Significant variables identified in the multi-level Poisson model are similar to those identified by previous studies (Winter et al., 2015; Witt & Green, 2018). One difference is the decreased RR associated with not separating ewes with FR. The separation of lame animals at the time of treatment was shown by Witt and Green (2018) to significantly decrease the RR of lameness in a clinical trial setting, while separating sporadically through the year was not effective. The questionnaire responses don’t include this differentiation, so this may impact the results. Another notable result is the decreased RR associated with increased soil clay content. The impact of soil clay content has only been assessed in a laboratory setting(Muzafar et al., 2016), so potentially there are other interactions in the natural environment that are unknown. The observed spatial autocorrelation in the model residuals suggests that the environmental factors included in the model are accounting for some, but not all, of the spatial variation in this data set. This model doesn’t include variable environmental factors, such as rainfall and air temperature, which have been observed to influence prevalence of lameness, specifically lameness caused by FR (Beveridge & Gregory, 1941; Graham & Egerton, 1968; Woolaston, 1993).

Figure 1
Image
Figure 2
Image
Figure 3
Image
Figure 4
Image

Acknowledgements

I would like to acknowledge my supervisors, Prof. Laura Green (University of Birmingham) and Prof. Matt Keeling (University of Warwick), and my funders at BBSRC through the Midlands Integrative Biosciences Training Partnership (MIBTP).

References

Anselin, L. (1995). Local Indicators of Spatial Association - Lisa. Geographical Analysis, 27(2), 93-115. doi:10.1111/j.1538-4632.1995.tb00338.x Beveridge, W. I. B., & Gregory, T. S. (1941). Foot-rot in sheep: a transmissible disease due to infection with Fusiformis nodosus (n. sp.); studies on its cause, epidemiology, and control. Melbourne,: H.E. Daw, government printer. Frosth, S., Konig, U., Nyman, A. K., Pringle, M., & Aspan, A. (2015). Characterisation of Dichelobacter nodosus and detection of Fusobacterium necrophorum and Treponema spp. in sheep with different clinical manifestations of footrot. Vet Microbiol, 179(1-2), 82-90. doi:10.1016/j.vetmic.2015.02.034 Graham, N. P., & Egerton, J. R. (1968). Pathogenesis of ovine foot-rot: the role of some environmental factors. Aust Vet J, 44(5), 235-240. Muzafar, M., Green, L. E., Calvo-Bado, L. A., Tichauer, E., King, H., James, P., & Wellington, E. M. (2016). Survival of the ovine footrot pathogen Dichelobacter nodosus in different soils. Anaerobe, 38, 81-87. doi:10.1016/j.anaerobe.2015.12.010 Roberts, D. S., & Egerton, J. R. (1969). The aetiology and pathogenesis of ovine foot-rot. II. The pathogenic association of Fusiformis nodosus and F. necrophorus. J Comp Pathol, 79(2), 217-227. Smith, E. M., Green, O. D., Calvo-Bado, L. A., Witcomb, L. A., Grogono-Thomas, R., Russell, C. L., . . . Green, L. E. (2014). Dynamics and impact of footrot and climate on hoof horn length in 50 ewes from one farm over a period of 10 months. Vet J, 201(3), 295-301. doi:10.1016/j.tvjl.2014.05.021 Winter, J. R., Kaler, J., Ferguson, E., KilBride, A. L., & Green, L. E. (2015). Changes in prevalence of, and risk factors for, lameness in random samples of English sheep flocks: 2004-2013. Prev Vet Med, 122(1-2), 121-128. doi:10.1016/j.prevetmed.2015.09.014 Witcomb, L. A., Green, L. E., Kaler, J., Ul-Hassan, A., Calvo-Bado, L. A., Medley, G. F., . . . Wellington, E. M. (2014). A longitudinal study of the role of Dichelobacter nodosus and Fusobacterium necrophorum load in initiation and severity of footrot in sheep. Prev Vet Med, 115(1-2), 48-55. doi:10.1016/j.prevetmed.2014.03.004 Witt, J., & Green, L. (2018). Development and assessment of management practices in a flock-specific lameness control plan; a stepped-wedge trial on 44 English sheep flocks. Prev Vet Med, Accepted for publication. Woolaston, R. R. (1993). Factors affecting the prevalence and severity of footrot in a merino flock selected for resistance to Haemonchus contortus. Aust Vet J, 70(10), 365-369. Youatt, W. (1869). Sheep: Their breeds, management, and diseases. London, UK: Simpkin Marshall & Co.

Keywords: Footrot, Lameness in sheep, Multi-level modeling, Multi-variable modeling, Spatial autocorrelation (SA), Local Moran's I

Conference: GeoVet 2019. Novel spatio-temporal approaches in the era of Big Data, Davis, United States, 8 Oct - 10 Oct, 2019.

Presentation Type: Student oral presentation

Topic: Spatial methods for environmental & exposure epidemiology and climate change

Citation: Witt JD (2019). Assessment of spatial, environmental, and management effects on lameness prevalence in UK sheep flocks. Front. Vet. Sci. Conference Abstract: GeoVet 2019. Novel spatio-temporal approaches in the era of Big Data. doi: 10.3389/conf.fvets.2019.05.00096

Copyright: The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers. They are made available through the Frontiers publishing platform as a service to conference organizers and presenters.

The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated.

Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed.

For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions.

Received: 31 May 2019; Published Online: 27 Sep 2019.

* Correspondence: Ms. Jessica D Witt, University of Warwick, Coventry, United Kingdom, red_331@hotmail.com