Introduction
Cystic echinococcosis (CE) is a helminthic zoonosis caused by the canid tapeworms, Echinococcus granulosus sensu lato. This disease occurs in a number of herbivorous and omnivorous animals (e.g. sheep, goat, cattle, camel, horse, buffalo, cervid and swine) as intermediate hosts and humans as accidental hosts by ingesting parasite eggs in contaminated water, soil, vegetables, etc. CE is globally distributed and is common in some regions of the Mediterranean Basin, South America, south-western and central Asia, Siberia, western China, northern Africa, southern and eastern Europe and the Middle East (Farhadi et al., Reference Farhadi, Haniloo, Rostamizadeh and Faghihzadeh2018; Gao et al., Reference Gao, Wang, Shi, Steverding, Wang, Yang and Zhou2018; Tamarozzi et al., Reference Tajbakhsh, Eisakhani and Fazl2018; Khademvatan et al., Reference Khademvatan, Majidiani, Foroutan, Tappeh, Aryamand and Khalkhali2019; Shafiei et al., Reference Sarkari, Sadjjadi, Beheshtian, Aghaee and Sedaghat2020). In fact, CE has been reported in a wide range of geo-climatic conditions such as circumpolar, temperate, subtropical and tropical zones (Rinaldi et al., Reference Rinaldi, Musella, Biggeri and Cringoli2014). Iran, as an endemic area in the Middle East with a variety of ecological and geographical conditions, is a good foci for CE development (Karamian et al., Reference Karamian, Haghighi, Hemmati, Taylor, Salehabadi and Ghatee2017). Similar to other parasites, E. granulosus biology and transmission are affected by the host–parasite association, behavioural traits, population density, habitat, environment and climatic conditions. For example, Echinococcus spp. ova survival can be affected by different geo-temporal parameters, including high temperature, low humidity and aridity (Veit et al., Reference Thevenet, Jensen, Drut, Cerrone, Grenóvero, Alvarez, Targovnik and Basualdo1995; Eckert & Deplazes, Reference Eckert and Deplazes2004; McManus, Reference McManus2010; Atkinson et al., Reference Atkinson, Gray, Clements, Barnes, McManus and Yang2013; Yan et al., Reference Veit, Bilger, Schad, Schäfer, Frank and Lucius2016; Byers et al., Reference Byers, Schmidt, Pappalardo, Haas and Stephens2019; Kołodziej-Sobocińska, Reference Kołodziej-Sobocińska2019). Understanding how parasites and their transmission are influenced by climatic changes and environmental conditions has become one of the most interesting topics among parasitology researchers (Okolo et al., Reference Okolo, Nnadi, Onyedibe and Ita2012). In recent years, the role of climatic changes and environmental conditions has been evaluated using specific tools and data analysis software programmes. Among the most valuable and functional tools are geographic information systems (GIS) and remote sensing technologies in health system programmes (Yilma & Malone, Reference Yan, Liang, Zheng and Zhu1998; Ghatee et al., Reference Ghatee, Sharifi, Haghdoost, Kanannejad, Taabody, Hatam and Abdollahipanah2013; Lyseen et al., Reference Lyseen, Nøhr, Sørensen, Gudes, Geraghty, Shaw, Bivona-Tellez and Group2014). GIS technology, as a computer-based system, analyses geospatial-referenced data and plays an important role in the surveillance of parasitic diseases. Researchers have used these approaches to study the interactions of parasites, hosts and reservoirs with geo-climatic factors, identify high- or low-risk areas for the disease spread and plan control strategies (Danson et al., Reference Danson, Bowyer, Pleydell and Craig2004; Rinaldi et al., Reference Rinaldi, Brunetti, Neumayr, Maestri, Goblirsch and Tamarozzi2006; Brundu et al., Reference Brundu, Piseddu, Stegel, Masu, Ledda and Masala2014). Although CE is a health problem in the southern and northern hemispheres (Eckert & Deplazes, Reference Eckert and Deplazes2004), most GIS-based studies have investigated the correlation between eco-geographical factors and the distribution of alveolar echinococcosis, and only a few studies have addressed the spatial distribution of CE using this approach (Staubach et al., Reference Sobhani, Zengir and Yazdani2001; Danson et al., Reference Danson, Bowyer, Pleydell and Craig2004; Pleydell et al., Reference Pleydell, Raoul, Tourneux, Danson, Graham, Craig and Giraudoux2004; Graham et al., Reference Graham, Danson and Craig2005; Cringoli et al., Reference Cringoli, Rinaldi, Musella, Veneziano, Maurelli, Di Pietro, Frisiello and Di Pietro2007). Therefore, in this study, we sought to determine the effects of geo-climatic factors on CE in south-western Iran (SWI) using the GIS approach during 2016–2018.
Materials and methods
Study area
SWI (fig. 1), with a surface area of about 230,567 km2, is located at approximately 27°N–31°N latitude and 47°E–55°E longitude and inhabits a population of 11.5 million people. The studied area consists of 74 counties in the four provinces of Khuzestan (27 counties), Fars (29 counties), Bushehr (ten counties), and Kohgiluyeh and Boyerahmad (eight counties). There are various climatic regions in SWI, including the Zagros mountain range extending from north to south, with moderately cold winters and temperate summers, plains with rainy mild winters and hot and dry summers and coastal regions of the Persian Gulf, with mild winters and hot and humid summers. These geo-climatic variations cause a variety of plants and land cover, and, consequently, various domestic and wild animals in SWI. The main native domestic animals in this area include sheep, goat, cattle, camel, buffalo, horse, donkey, dog and cat, and wild animals are deer, wild goat, ram, ewe, fox, wolf, jackal, bear and panther (Ghasem et al., Reference Ghasem, Shamsipour, Miri and Safarrad2012; Sagheb-Talebi et al., Reference Rinaldi, Musella, Biggeri and Cringoli2014). According to the Veterinary Bureau consensus, the population of small and large livestock was over 13 million in these provinces during 2016–2018. In detail, it was about 6.2, 4.4, 2.2 and 1 million in Khuzestan, Fars, Kohgiluyeh and Boyerahmad, and Bushehr provinces, respectively. These areas are dominated by different land covers, ranging from forests and irrigated and rich agricultural lands to bare and salty areas. Altitude varies from Dena Mountain (4283 m, as the second highest summit of Iran) in the Zagros Mountains to the lowest grounds in Khuzestan and Bushehr provinces, where there are tens of hundreds of kilometres of coastal areas alongside the Persian Gulf. Various urban, rural and nomadic populations live in SWI. Nomadic tribes live on herding and migrate annually from warm areas to cold ones – that is, from highland pastures in summer (Yailaq) to winter habitations (Qishlag) and vice versa, with their livestock and shepherd dogs. These animals have been found to play a role in the prevalence of some parasitological diseases in SWI (Sarkari et al., Reference Sagheb-Talebi, Pourhashemi and Sajedi2010; Ghatee et al., Reference Ghatee, Haghdoost, Kooreshnia, Kanannejad, Parisaie, Karamian and Moshfe2018; Kanannejad et al., Reference Kanannejad, Haghdoost, Ghatee, Azarifar, Shahriari and Moshfe2019; Moshfe et al., Reference Moshfe, Sarkari, Arefkhah, Nikbakht, Shahriarirad, Rezaei, Jamshidi and Moradian2019; Ghadimi-Moghadam et al., Reference Ghadimi-Moghadam, Salahi, Ghatee, Ghadimi-Moghadam, Kanannejad, Mosavi, Ramshk and Khoramrooz2020).
Livestock CE data collection
Comprehensive information about slaughtered livestock (e.g. number and type of animals, infection type, infected tissue, carcass weight, etc.) is routinely recorded by the Iranian National Veterinary Organization. Therefore, we retrieved the required data, including the total number and types of livestock (i.e. goat, sheep, cattle, camel and buffalo) infected with CE, from the veterinary bureaus of provinces of SWI for a three-year period (2016–2018). The data were categorized based on the livestock species belonging to each county and then entered into Microsoft Excel, (2016).
Geographical and climatic data collection
Meteorological data included mean annual rainfall (MAR), mean annual humidity (MAH), mean annual evaporation (MAE), the annual frequency of frost days, the mean number of sunny hours, mean annual wind speed, mean annual temperature (MAT), minimum MAT (MinMAT) and maximum MAT (MaxMAT). The data for the study period were retrieved from all the 62 synoptic stations (stations that collect meteorological data every 3 or 6 h) located in the studied areas. MAR, MAH and MAE raster layers were developed by the kriging interpolation method, and other meteorological layers were generated by the tension-based spline interpolation method with a resolution grid of 2 × 2 km after examining various interpolation models. Digital elevation model (DEM) raster layers and land cover vector layers (including features covering the province's surface) were retrieved from the Provincial Department of Agricultural Affairs. Layers belonging to the four provinces were merged to obtain one layer including all the SWI areas for each mentioned map. The slope raster layer was generated based on the DEM map by calculating the maximum rate of change in value between each cell and its neighbours by using the spatial analyst tool.
Geographical and climatic analysis
All the geo-climatic data were analysed using ArcGIS version 10.5 (http://www.esri.com/arcgis). The shapefile polygon layer of the SWI counties whose rate of total infection was previously enrolled within its attribute was used as the basic layer. The geometric intersection of the counties layer and the land cover polygonal shapefile was computed by an identity tool and then dissolved to unify all land cover feature polygons for each county. The sum of area (square km) of each land cover feature was calculated using the geometry tool for each county. Mean pixel values of DEM, slope, MAR, MAT, MaxMAT, MinMAT, MAH, MAE, frost days and sunny hours’ raster layers were calculated for each county by zonal statistics in the related tables. A final table encompassing all of the aforementioned data for SWI counties was drawn, and then it was converted to Excel format for statistical analysis.
Statistical analysis
Having described the geographical distribution of livestock CE in SWI, in the first step of statistical analysis, we investigated the correlation between the total infected livestock rate (dependent variable) and geo-climatic factors (independent variables). Then, since all the variables in the study were quantitative, a stepwise linear regression model was developed for variables that had a significant correlation with the dependent factor to investigate the probable predictors. The analyses were carried out using SPSS, version 21 (IBM Corp).
Results
Geographical distribution of livestock CE in SWI
The overall prevalence of CE in slaughtered livestock was 9% in SWI. The highest and lowest rates of infection belonged to buffalo (11.6%) and goat (7.7%). Also, the overall highest infection rate was related to Kohgiluyeh and Boyerahmad (14.3%) and the lowest infection rate was found in Fars province (7.6%) (table 1). Based on the geographical distribution of CE, the most infected counties were distributed throughout a hypothetical band stretching from north-west (north and centre of Khuzestan province) towards central and south-eastern areas (i.e. from Kohgiluyeh and Boyerahmad to the centre of Fars province). The distribution of infected livestock is shown in fig. 2(a–f).
– There is no camel and buffalo husbandry.
Effects of geographical and climatic factors on CE
Pearson correlation analysis showed that MAR, frost days, elevation, slope and surface of semi-condensed forest landcover were positively associated with CE rate in SWI, and rainfall was found to have the highest correlation. MAE, MAT, MaxMAT, MinMAT and salinity land cover showed a significant negative correlation (P < 0.05) with the CE occurrence (table 2 and fig. 3). A decreasing trend in CE rate was found with the increase in sunny hours in most parts of SWI (areas with 3358–3652 sunny hours based on the map); however, no significant correlation was found in this regard (P = 0.09; fig. 3). Also, there was no significant correlation between humidity, wind speed and land covers of bare lands, thin rangelands, condensed rangelands, irrigated farms and gardens, semi-condensed rangelands, riverbeds, swamps, urban, sandy areas and canebrakes with CE infection rate, although an inverse R was recorded for them. Positive but not significant correlations were shown between the distribution of CE and condensed forest rained farms and gardens, and thin forest land covers (table 2). To evaluate the effect of these factors on the occurrence of CE, a stepwise linear regression model was developed with the model fitness of approximately 60% (R square = 0.601). MAE (P < 0.001; beta = −0.60) was found to be a predictive factor in the stepwise model, while MAT (P = 0.841; beta = −0.024), MaxMAT (P = 0.85; beta = −0.02), MinMAT (P = 0.94; beta = −0.01), MAR (P = 0.67; beta = 0.07), elevation (P = 0.89; beta = −0.01), slope (P = 0.65; beta = −0.05), frost days (P = 0.84; beta = 0.02), semi-condensed forests (P = 0.72; beta = −0.04) and salinity land covers (P = 0.54; beta = −0.06) were excluded from the model.
* Significant factors (P-value < 0.05).
– Negative correlation with CE occurrence.
MAR, mean annual rainfall; MaxMAT, maximum mean annual temperature; MinMAT, minimum mean annual temperature; MAT, mean annual temperature.
Discussion
The current study revealed that MAE is the most important predictor of livestock CE in SWI, although CE infection rate was also positively correlated with MAR, frost days, elevation, slope and surface of semi-condensed forest land cover, and negatively with MAT, MaxMAT, MinMAT and salinity land cover. There was a trend between sunny hours and CE occurrence in this region.
The effect of MAE as the most important predictor of CE distribution in SWI is probably due to its impact on soil aridity that decreases egg viability and transmission capability. Some studies confirmed that Echinococcus spp. egg is sensitive to dryness and survives only for a short time if exposed to direct sunlight and dryness (Thevenet et al., Reference Tamarozzi, Akhan and Cretu2005). The study by Staubach et al. (Reference Sobhani, Zengir and Yazdani2001) showed dryness as a limiting factor in the tenacity of Echinococcus multilocularis oncospheres in eggs. Furthermore, increased evaporation is related to the expansion of drought and salt lands, and reduced growth of plants and rangelands, which can provide a hostile environment for egg survival (Tajbakhsh et al., Reference Staubach, Thulke, Tackmann, Hugh-Jones and Conraths2015).
Our study revealed an inverse relationship between salinity and CE occurrence in SWI, explaining why husbandry in this area is less common than elsewhere. On the other hand, due to the presence of salt in this area and its ovicidal effect, the parasite's life cycle will most likely not be completed (Craig & Macpherson, Reference Craig and Macpherson1988)
Semi-condensed forests directly affect CE occurrence in SWI. In recent years, climatic changes and drought have increased in the Middle East, especially in Iran (as a country with more than 82% of its area locates in arid and semi-arid zones), where the average rainfall (250 mm) has decreased to one-third of the world's rainfall average (860 mm) (Amiri & Eslamian, Reference Amiri and Eslamian2010; Sobhani et al., Reference Shafiei, Ghatee, Jafarzadeh, Javanshir and Karamian2020). Drought is more severe in some parts of Iran such as south-western, southern, south-eastern and central regions (Raziei et al., Reference Raziei, Saghafian, Paulo, Pereira and Bordi2009; Abarghouei et al., Reference Abarghouei, Zarch, Dastorani, Kousari and Zarch2011). Climate change and drought have forced rural and nomadic communities to migrate and settle in greener places such as semi-condensed forest areas, river banks and green rangeland areas, and wherever water sources and feedstuff are available for their livestock (Keshavarz et al., Reference Keshavarz, Karami and Vanclay2013; Rashednasab et al., Reference Rashednasab, Ahmadvand and Sharifzadeh2018). Higher soil wetness and more shadow in these regions may increase the chance of survival of eggs excreted by the definitive host, leading to the infection of intermediate hosts. Also, semi-condensed forests, as the habitat of carnivorous canids such as stray dogs, wolves, foxes and jackals, as the definitive hosts for E. granulosus, are mostly located at higher-altitude hillsides and in the close proximity of mountainous villages and nomadic shelters, which are suitable conditions for the maintenance of the parasite cycle (Otero-Abad & Torgerson, Reference Otero-Abad and Torgerson2013; Farimani et al., Reference Farimani, Raufirad, Hunter and Lebailly2017).
Other significant factors related to CE in the present study were rainfall, temperature, elevation, slope and frost days. According to fig. 3, there is a comprehensive overlap between these factors. For example, with increasing levels of elevation, elevated MAR and disease burden have been reported. The effect of MAR on CE distribution is understandable. Rainfall contributes to the expansion of green rangelands and forests, and increases the amount of forage for livestock, which results in higher herding and infection rates. In addition, rainfall can increase the relative humidity and the balance between temperature and humidity, which is needed for the survival of parasite eggs. There was a negative association between different temperature models and CE in our study, and the schematic temperature map (fig. 3) showed that higher rates of CE infection have been found in areas with 15°C–29°C MAT, while CE prevalence was reduced in areas where MAT was above 30°C. This is because of higher susceptibility of parasite eggs to higher temperatures, dryness and dehydration. For example, in some studies it was observed that Echinococcus eggs were disinfected by high temperatures (hot water of 85°C or above) and desiccation, while they could survive in freezing conditions. However, they can be killed by freezing at –80°C for 48 h or –70°C for four days (Colli & Williams, Reference Colli and Williams1972; Eckert & Deplazes, Reference Eckert and Deplazes2004; Federer et al., Reference Federer, Armua-Fernandez, Hoby, Wenker and Deplazes2015; Cédric et al., Reference Cédric, Frits and Silivia2019). The results of the Harriott et al. (Reference Harriott, Gentle, Traub, Cobbold and Magalhães2019) study in Queensland, Australia, showed that CE infection rate significantly increased with higher rainfall, relative humidity and temperature. Veit et al. (Reference Thevenet, Jensen, Drut, Cerrone, Grenóvero, Alvarez, Targovnik and Basualdo1995) presented that E. multilocularis eggs were highly sensitive to elevated temperatures and to desiccation as far as the eggs’ infectivity was lost after 3 h at 45°C. Kern et al. (Reference Kern, Ammon, Kron, Sinn, Sander, Petersen, Gaus and Kern2004) presented that E. multilocularis eggs can remain infective for months under suitable conditions such as low temperatures and high humidity.
Although in the present study there was no significant correlation between MAH and CE prevalence, most counties with higher rates of infection were within 35–40% relative humidity range. In some counties of Bushehr and Khuzestan provinces with high humidity (>40%), the prevalence of CE was lower than other counties in SWI. This finding can be explained by the higher MAT, MaxMAT and MAE in these areas, which can reduce the infection rate.
It should be noted that due to the lack of accurate information as to the geographical distribution of the parasite's definitive and accidental hosts (carnivores and human) in SWI during the present study, further comprehensive studies are recommended to determine the effects of geo-climatic factors on the prevalence of CE in the parasite's definitive and intermediate hosts in SWI and other regions of Iran.
Conclusion
In short, GIS-based analysis showed that some of the geo-climatic factors such as MAE, as the main predictive factor, and MAT, MinMAT, MaxMAT, MAR, elevation, slope, land cover of the semi-condensed forest and salinity lands had significant correlation with CE distribution in SWI. Accordingly, areas with lower evaporation and temperature and higher rainfall, especially those covered by semi-condensed forest, mostly found in mountainous higher altitudes, should be prioritized for consideration as probable CE risk zones. GIS, as a powerful descriptive and analytical tool that identifies risk factors associated with livestock population and the distribution of hydatidosis, can help health professionals and veterinarians identify high-risk regions and assess the performance of control programmes in SWI.
Acknowledgements
This article is part of a PhD thesis in medical parasitology (number A-12-95-13) supported by Zanjan University of Medical Sciences. Special thanks and appreciation to the Veterinary Bureau consensus and Weather Bureau in Fars, Khuzestan, Kohgiluyeh and Boyerahmad, and Bushehr provinces for their assistance in collecting meteorological, geo-climatic data and information of livestock's CE in SWI required for this study. We also thank all collaborators involved in the project.
Financial support
This article was financially supported by Zanjan University of Medical Sciences (No: A-12-95-13) and approved by Ethics committee of Zanjan Univwersity of Medical Sciences (Code: IR.ZUMS.REC.1397.252).
Conflicts of interest
None.