Predictive Analytics of Cattle Host and Environmental and Micro-Climate Factors for Tick Distribution and Abundance at the Livestock-Wildlife Interface in the Lower Okavango Delta of Botswana.

Tick distribution and abundance is influenced by several factors that include micro-climate and environmental and host factors. Contextual understanding of the role played by these factors is critical to guide control measures. The aim of this study was to determine the predictive value of these factors for changes in tick distribution and abundance at the livestock-wildlife interface of the lower Okavango Delta. A two-stage quota sampling design was adopted to select 30 clusters of seven cattle each for tick investigation. Tick investigation was done by lifting the tail to count the total number of ticks at the anno-vulva region during the four meteorological seasons of the year. Additional data were collected on ear tag number, location of origin, sex, age, body condition score (BCS), season of the year, stocking density, and Normalized Difference Vegetation Index values of source terrain. A random effects model was used to evaluate predictive value of the above for tick abundance. Additional mapping of tick distribution pattern in the abattoir catchment area was conducted using spatial autocorrelation and hot-spot analysis. Tick intensity of infection increased linearly from males to females (Z = 3.84, p < 0.001), decreased linearly from lower to higher BCS (Z = -4.11, p < 0.001), and increased linearly from cold-dry through dry to wet seasons (Z = 10.19, p < 0.001). Significant clustering of neighboring crushes on account of tick count was noted in the late-hot-dry season, with high tick count in crushes located along the seasonal flood plains and low tick counts in those located in the dry grasslands. It was concluded from this study that cattle tick abundance is influenced largely by season of the year and that the micro-climatic conditions brought about by the seasonal flooding of the delta have a decided effect on tick distribution during the driest of the seasons. Our study has, for the first time, profiled drivers of tick distribution and population growth in this unique ecosystem. This has the potential to benefit human and veterinary public health in the area through implementation of sustainable tick control strategies that are not heavily reliant on acaricides.

Rumen impaction in cattle associated with ingestion of the pupal cocoons of Gonometa spp. in Botswana.

Mortality in cattle associated with ingestion of cocoons (matlhoa in Setswana) of both Gonometa postica and Gonometa rufobrunnea is rare and has only previously been reported in South Africa, Zimbabwe and Namibia. A case history of gradual weight loss, bloat, dyschezia with dry faeces and laboured gait, resulting in sudden death after drinking water and associated with ingestion of pupal cocoons of Gonometa spp., was reported by keepers at Mmaditau crush in Botswana in 2013. The crush was a shared holding in a communal area with 15 registered animal keepers. The objective of this study was to profile the history, clinical signs, post-mortem findings, morbidity and mortality from the outbreak using the descriptive study method. Altogether, 81 cattle out of a total of 507 died of impaction from August to December 2013. On autopsy, a loosely connected mass of ingesta, intertwined in ropy silky strands, was observed. It was concluded that there is no readily accessible and available form of treatment at crush level, leaving only evasive husbandry practices as the feasible option. To aid evasive husbandry management practices, temporal and spatial monitoring of population dynamics of Gonometa spp. is recommended, particularly during a drought spell when animals are prone to develop pica, as the basis for an early warning system to farmers.

An elaborated SIR model for haemonchosis in sheep in South Africa under a targeted selective anthelmintic treatment regime.

Infection with the abomasal nematode Haemonchus contortus is responsible for considerable production loss in small ruminants globally, and especially in warm, summer-rainfall regions. Previous attempts to predict infection levels have followed the traditional framework for macroparasite models, i.e. tracking parasite population sizes as a function of host and climatic factors. Targeted treatment strategies, in which patho-physiological indices are used to identify the individuals most affected by parasites, could provide a foundation for alternative, incidence-based epidemiological models. In this paper, an elaboration of the classic susceptible-infected-recovered (SIR) model framework for microparasites was adapted to haemonchosis and used to predict disease in Merino sheep on a commercial farm in South Africa. Incidence was monitored over a single grazing season using the FAMACHA scoring system for conjunctival mucosal coloration, which indicates high burdens of H. contortus, and used to fit the model by estimating transmission parameters. The model predicted force of infection (FOI) between sequential FAMACHA monitoring events in groups of dry, pregnant and lactating ewes, and related FOI to factors including climate (temperature, rainfall and rainfall entropy), using a random effects model with reproductive status group as the cluster variable. Temperature and rainfall in the seven days prior to monitoring significantly predicted the interval FOI (p≤0.002), while rainfall entropy did not (p=0.289). Differences across the three groups accounted for approximately 90% of the variability in the interval FOI over the period of investigation. Maintained FOI during targeted treatment of cases of haemonchosis suggests strong underlying transmission from sub-clinically infected animals, and/or limited impact on pre-existing pasture contamination by removal of clinical worm burdens later in the grazing season. The model has the potential to contribute to the sustainable management of H. contortus by predicting periods of heightened risk, and hence to focus and optimise limited resources for monitoring and treatment. SIR-type model frameworks are an alternative to classic abundance-based compartmental models of macroparasite epidemiology, and could be useful where incidence data are available. Significant challenges remain, however, in the ability to calibrate such models to field data at spatial and temporal scales that are useful for decision support at farm level.