Background
Anthrax is a zoonotic disease of concern that occurs naturally in herbivorous wildlife and livestock thereby significantly affecting human livelihoods and biodiversity. The disease is one of the neglected tropical diseases which is caused by the gram-positive spore-forming bacterium
Bacillus anthracis [
1]
. In terms of geographic distribution and endemism
, anthrax is found in several regions across the globe such as Asia, Australia, North and South America, Southern parts of Europe, sub-Saharan Africa and Central and South America [
2,
3]. The spatial distribution of the disease is attributed to the ability of
B. anthracis to form spores that thrive well under diverse environmental conditions [
4]. Although the disease burden of anthrax in herbivores is not fully known, studies have shown that anthrax outbreaks occur almost every year killing thousands of animals and transmitting the disease to humans upon consumption of the meat [
1]. The disease is of global concern as it results in high animal mortality with subsequent threats to human health [
5,
6]. Despite a decrease in reported livestock anthrax cases globally in the past decade [
7] between 20,000 to 100,000 cases of the disease are still being recorded each year especially in developing countries [
4]. The disease also affects human beings with 1.83 billion people living within high anthrax-risk areas and Africa recording the highest human incidences of the disease [
4]. In fact, human anthrax cases often associated with animal anthrax epidemics in resource poor communities occur at least every year in African countries such as Zambia, Zimbabwe, and Ethiopia [
8,
9]. Thus, there is need to develop or adopt methods that allow for better understanding of current and future spatial distribution of anthrax as a preamble to identifying potential anthrax hotspots [
10].
Zimbabwe has an estimated cattle herd of ~ 5.5 million with 90% of the national cattle herd under the smallholder sector [
11]. Over the years, the national herd has declined due to increased mortality from anthrax and tick-borne diseases such as January disease and Heart water [
12]. Previous studies have reported the following cattle deaths emanating from tick-borne diseases in Zimbabwe: 3,430 in 2017; 1,133 in 2018; 1,903 in 2019; 2,772 in 2020 and 1,478 died in 2021 [
13]. In fact, it has been reported that cattle deaths from tick-borne diseases can be as high as 9% of the national heard [
14]. This is despite an increase in the surveillance and disease control measures to curtail the occurrence and spread of tick-borne diseases by the Department of Veterinary Services [
14]. Typical anthrax outbreaks in the country are usually recorded during the dry (July to October) and wet (November to February) seasons.
Anthrax is transmitted via several modes in livestock and wildlife. Ingestion of spores during grazing in landscapes that previously experienced anthrax outbreaks is the primary mode of transmission in animals [
15,
16]. Scavenging animals, biting flies or poor disposal of infected animal carcasses facilitate disease transmission through exposing vegetative cells to oxygen thereby resulting in spore formation [
17]. Vaccination and proper carcass disposal are the main methods of control in the event of outbreaks.
The spatial distribution of anthrax is influenced by several factors which include livestock density, soil pH, availability of surface water, rainfall, temperature dynamics and vegetation cover [
18]. High livestock density increases interaction among individual animals thereby increasing anthrax transmission [
19]. The interaction usually occurs as livestock forage for resources including when searching and drinking surface water. The interaction is especially intense during the dry season when there are limited waterholes thereby resulting in increased interaction as livestock from different geographical regions mix unlike in the wet season when water sources are ubiquitous [
20]. During the dry season when pastures are scarce and the grass has become shorter, there is a high probability for animals to consume the grass together with soil particles often leading to abrasions in the mouth thus increasing chances of disease transmission in contaminated areas. On the other hand, soils which are slightly alkaline (pH of 6.74) and contains high calcium levels help to maintain the
B. anthracis spore cell wall integrity. This results in continued persistence of anthrax in endemic areas. Of late, climate change seems to be a key driver influencing anthrax occurrence and distribution [
21]. Heavy rains and floods following a long dry period combined with high temperature results in transportation and deposition of
B. anthracis spores in low-lying areas as well as speeding up the bacterium life cycle [
22].
Numerous studies covering different aspects of anthrax have been carried out at different spatial and temporal scales in Zimbabwe. These studies include those that assessed the ecological niche of
B. anthracis [
23,
24] and those that focused on spatial and temporal distribution of anthrax [
9,
25], anthrax in animals [
26‐
28] and humans [
29‐
39]. Studies have also assessed influence of politics on anthrax control [
40] as well as its impact on rural livelihoods [
9]. Although these studies have improved the understanding of anthrax ecology, spread and dynamics in both space and time, they lack futuristic insights into the potential effects of climate change on anthrax occurrence in Zimbabwe. Information on the distribution of anthrax is important for anthrax control and management strategies, such as the targeted vaccinations, optimizing resource allocation and prioritisation of prevention and control strategies in high-risk areas [
41]. This is particularly important in a resource-poor country such as Zimbabwe, where the anthrax vaccines are often inadequate to cover all livestock across the country. Therefore, the objectives of this study were to determine the current distribution of anthrax outbreaks as well as predict the future habitat suitability and distribution of anthrax occurrence using bioclimatic predictors. This is important to inform surveillance, control and prevention strategies which need to be undertaken by veterinary and public health personnel.
Discussion
Results of this study predicted highly suitable areas for anthrax outbreaks in the western and eastern parts of Zimbabwe. The current suitability map shows an increase in highly suitable areas of anthrax compared to previous studies. These results suggest an increase in bioclimatically suitable areas for the disease as well as the superiority of ensemble modelling that integrated eight species distribution models over a single species distribution model (MAXENT), e.g., [
25]. In contrast, the study predicted that the northern parts of the country would remain marginally suitable, suggesting that these areas could be less likely to experience anthrax outbreaks and therefore may require less attention relative to other districts. However, the overall results suggest a variable increase in future distribution of anthrax occurrence thereby requiring monitoring of the disease to reduce its impacts [
33]. Vaccinations are still one of the best methods to control anthrax and livestock should be vaccinated annually to reduce the incidence of the disease [
9]. The first step in implementing vaccination is to determine the priority areas to target hence maps generated in this study can be used to for targeted surveillance and vaccination in the country factoring in their different challenges [
63]. Therefore, resources could be channelled towards areas that are projected to be suitable for anthrax in the country [
64].
Although this study used cattle anthrax outbreaks only, the models were able to predict wildlife areas such as Hwange National Park as suitable for the disease. However, it is well known that the entire periphery and the interior of a wildlife area is usually shared by livestock and wild animals, and hence the possibility of increased anthrax transmission [
65]. This means that the distribution might be expanded if there is spatial overlap between wildlife and livestock which is a common phenomenon at wildlife-livestock interfaces. The areas close to wildlife were predicted to be highly suitable in the future and thus need close monitoring and strategic vaccinations to prevent and reduce the likely future anthrax outbreaks [
66]. Previous research in Kenya predicted anthrax occurrence in entire wildlife sanctuaries such as Nakuru National Park [
67]. Similarly, the present forecast maps predicted anthrax occurrence in entire Zimbabwean wildlife areas.
From this study, the occurrence and distribution of anthrax was observed to be related to various climate variables. For example, precipitation of warmest quarter (Bio18), minimum temperature of the coldest month (Bio6) and precipitation seasonality (Bio15) were more important in modelling the current distribution of anthrax. Similar findings were observed in western Uganda and Western Africa where seasonality of precipitation and temperature in the warmest months were found to affect the distribution of anthrax [
1,
44]. Previous studies by Chikerema et al. [
25] showed an increased anthrax outbreak occurrence during the hot dry months in Zimbabwe. This supports findings of this study where precipitation of the warmest quarter contributed more to disease occurrence. Districts such as Beitbridge, Gwanda, Mwenezi, Chiredzi and Kariba were found to be marginally suitable for anthrax occurrence in both the current and future models. This might be due to low precipitation in these districts. Chikerema et al. [
25] reported rainfall as a contributing factor for the temporal and spatial occurrence of anthrax in Zimbabwe. The moisture provided by precipitation influences anthrax occurrence through exposing buried spores, collecting and concentrating spores in storage areas, and dispersing spores through runoff. The duration of the dry season is also related to anthrax occurrence. In addition, animals that graze short grasses close to the ground during the dry season are more likely to be exposed to spores thereby increasing the possibility of anthrax outbreaks. The dry season also results in water and forage shortages leading to a higher anthrax transmission in livestock and wildlife at remaining water points [
6].
Unlike previous studies, this study used ensemble modelling to assess the potential effects of climate change on the spatial distribution of anthrax. An Ensemble of eight different SDMs was used to understand distribution of anthrax which has an advantage of reducing omission and commission errors since all the prediction of the eight different models were taken into consideration [
68]. Ensemble modelling improves model performance resulting in better accuracy compared with a single predictive model [
69]. This is achieved through reducing the variance component of the prediction error. An ensemble model can make better predictions and perform better than any contributing model [
70]. Another important benefit of the ensemble method is a more robust or reliable average performance of the model. Robustness and reliability are the main concerns in machine learning projects. An ensemble reduces the spread or dispersion of the predictions [
71].
This study took into consideration bioclimatic parameters and elevation in modelling anthrax occurrence distribution in Zimbabwe. Future studies should include other environmental factors such as livestock density, soil pH and type, vegetation cover and type and water sources distribution to determine their influence on anthrax occurrence. Factors such as soil pH and type influence the survival of anthrax spores [
72] and earlier studies in the country identified soil type as an important predictor followed by variance of vegetation biomass and maximum temperature [
24]. Furthermore, the occurrence of anthrax in endemic areas is usually associated with pasture degeneration caused by over utilization or drought. The condition of the pastures leads to nutritional stress and herbivores are forced to feed on heavily utilized short grass or herbs and thereby contract anthrax through ingesting soil containing the spores. Hence, the height of the grass in the grasslands and wooded areas has an influence on the occurrence of anthrax outbreaks. On the other hand, burning is regarded as one of the preferred method of anthrax control through their extermination of most spores as was demonstrated during the massive wildlife anthrax outbreak in Zimbabwe [
27]. The method was used to disinfect the soil and vegetation and thereby avoid animals using areas of potentially high contamination and the same approach has been used in the Kruger National Park to sanitize the environment leading to a rapid decrease in the number of deaths due to anthrax outbreaks. Hence, it is likely that annual and extensive bush or grassfires might have an influence on the occurrence of anthrax. These environmental variables are important in the transmission dynamics of the disease and their inclusion in future studies may provide a more accurate potential predicted distribution. Although information on anthrax cases in human, livestock and wildlife is critical for achieving ONE Health, the information is not readily available in a consolidated format and would be critical for informing anthrax management policies and interventions.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.