Edaphic and climatic factors influence on the distribution of soil transmitted helminths in Kogi East, Nigeria

The need for a reliable risk map in the control of soil-transmitted helminths (STHs) in Kogi East, North Central Nigeria is very important. This study was carried out to determine the effect of environmental risk factors on geospatial distribution of STHs. Epidemiological data were obtained from a district-wide survey conducted in 2018 in Kogi East. Edaphic and climatic factors were downloaded as spatial layers from international recognised health data resources centres. A total of 24 environmental factors were used in determining the risk map of STHs using MaxEnt tool. The predicted high-risk areas of A. lumbricoides, hookworms and S. stercoralis were the central part of Kogi East covering parts of Dekina, Ofu, Igalamela-Odolu, Olamaboro and Omala LGAs with probability of 0.8 to 1.00. Among the factors investigated; Temperature [mean diurnal temperature range (BIO2), temperature annual range (BIO7) and maximum temperature of the warmest month (BIO5)], precipitation [precipitation of the wettest quarter (BIO16)], and soil clay contents were the five factors that exerted most significant influence on the geospatial distribution of STHs in Kogi East, Nigeria. Public health control programmes on STHs should target high-risk areas by including them in mass drug administration, health education as well as provision of water, sanitation and hygiene infrastructures.

Source of epidemiological data. The epidemiological data used for this study were obtained from an earlier district-wide survey carried out in 2018 (Table 1) 25 in rural communities of Kogi East, Kogi State, Nigeria. The study obtained samples from school-children of age 5 to 14 years. Samples collected were examined using formal ether sedimentation technique. The study was carried out in schools that did not receive anthelminthic drugs during the yearly periodic deworming exercise carried out by the State Ministry of Health. During the survey, the geographical coordinates of each school and community were captured within the school premises using a handheld Global Positioning system (GPS) device, Garmin 12XL (Garmin Corp, USA). www.nature.com/scientificreports/ Environmental data collection. Climatic and elevation variables. Remotely sensed environmental data for altitude, temperature and precipitation were obtained from Worldclim database 30 . The climatic variables such as temperature and precipitation are at global and meso scales and topographic variables such as elevation and aspect likely affect species distributions at meso and topo-scales 31 . Hence, the use of the climatic and topographic variables in the prediction of distributions of soil transmitted helminths in Kogi East, Nigeria. Also, temperature In this study, a total of 19 bioclimatic factors of present climate for Nigeria were downloaded at 1 km spatial resolution ( Table 2) from Worldclim database 30 and were used in the prediction of soil transmitted helminths distribution in Kogi East. Elevation data derived from the Shuttle Radar Topography Mission (SRTM) (aggregated to 30 arc-seconds, "1 km") were also downloaded from WorldClim database 30 .
Edaphic variable. The influence of edaphic factors on the distribution of STHs have been reported by several researchers globally [34][35][36] as important factors in the biology of STH parasites. In view of this, data for soil pH, soil moisture content, soil organic carbon and soil clay content for Africa continent were downloaded from International Soil Reference Centre (ISRIC) soil database as spatial layers (Table 2) 37 .
File conversions and resampling. The 19 bioclimatic factors downloaded from WorldClim data are in geographic coordinates of latitudes and longitudes which comes as .bil files were extracted into a folder. These data were transformed into predefined geographic coordinate system (GCS_WGS_1984), this projection was done on ArcMap 10.1 and were converted to asci files on DIVA-GIS 7.5. These files were transferred back to Arc-Map and assigned a projected coordinate system of Universal Transverse Mercator (UTM) Zone 32 N (Nigeria is located on UTM Zone 31, 32 and 33). Also, the edaphic factors obtained were also assigned a projected coordinate system. The projected raster files (i.e. climatic, elevation and edaphic) were all clipped into a layer using the administrative boundary map of the study area, this was downloaded on DIVA-GIS database 38 .
Prior to modelling, all variables were resampled from their native resolution to a common resolution of 1 km spatial resolution using the nearest neighbour technique on ArcMap 10.1 to enable overlaying of variables. The resampled raster files were converted to float files on ArcMap 10.1 and transferred to DIVA-GIS 7.5. Float files were converted to grid files and then to asci files on DIVA-GIS 7.5 and were used on MaxEnt tool for modelling the distribution of STHs in Kogi East.
Ecological niche modelling. The potential distribution of STHs were modelled using maximum entropy (MaxEnt) software version 3.3.3k 39 . MaxEnt uses environmental data at occurrence and background locations to predict the distribution of a species across a landscape 31,40 . This modelling tool was selected based on the reasons of Sarma et al. 41 , they stated that this tool allows the use of presence only datasets and model robustness is hardly influenced by small sample sizes. It has been shown to be one of the top performing modelling tools 42 . www.nature.com/scientificreports/ Probability of presence of each of the STH was estimated by MaxEnt using the prevalence of each of the STH parasites obtained for 45 sampled communities in the 9 LGAs of Kogi East during the district-wide survey carried out in 2018 25 served as the presence records to generate background points were used 41 . Regularization of the prevalence was performed to control over-fitting. This modelling tool uses five different features to perform its statistics; linear, quadratic, product, threshold and hinge features to produce a geographical distribution of species within a define area. The MaxEnt produces a logistic output format used in the production of a continuous map that provides a visualization with an estimated probability of species between 0 and 1. This map distinguish areas of high and low risk for STH infections 41 .
The 19 bioclimatic factors, elevation data and the edaphic factors obtained were used for the ecological niche modelling. The level of significance of contribution of the altitude and 19 bioclimatic factors was used to calculate the area under the receiver operating characteristics curve (AUC) was used to evaluate the model performance.
The AUC values varies from 0.5 to 1.0; an AUC value of 0.5 indicates that model predictions are not better than random, values < 0.5 are worse than random, 0.5-0.7 signifies poor performance, 0.7-0.9 signifies reasonable/ moderate performance and > 0.9 indicates high model performance 43 .
Model validation was performed as follows 41 , using the 'sub-sampling' procedure in MaxEnt. 75% of the parasites prevalence data were used for model calibration and the remaining 25% for model validation. Ten replicates were run and average AUC values for training and test datasets were calculated. Maximum iterations were set at 5000. Sensitivity, which is also named the true positive rate, can measure the ability to correctly identify areas infected. Its value equals the rate of true positive and the sum value of true positive and false negative. Specificity, which is also named the true negative rate, can measure the ability to correctly identify areas uninfected. Its value equals the rate of true negative and the sum value of false positive and true negative. LGA are areas that fell within the low risk areas (Fig. 3b). Probability ranges from 0.0 to 0.3 for low risk areas and 0.7-1.00 for high risk areas.
The high risk areas for hookworms were the same for S. stercoralis as well as the low risk areas (Fig. 3c). The combined predicted map for the STHs (Fig. 3d) indicates high risk areas in Dekina, Igalamela-Odolu, Idah, Ofu and Olamaboro LGAs with patches in Omala and Ibaji LGAs.
Model performance and influencing factors. The mean percent contribution and permutation importance are two factors used to assess the effectiveness of variables used in the modelling of STH in Kogi East. A total of 24 factors were assessed, categorized into elevation (altitude), climatic and edaphic variables. Of these variables assessed, mean diurnal range (BIO2) was the factor with the highest importance and contribution in determining the distribution of STHs in Kogi East. It had the highest PC and PI of 36.7% and 59.9%, 39.7% and 64.2%, and 36.5% and 57.5% for A. lumbricoides, hookworms and S. stercoralis respectively ( Table 3).
The receiver operating characteristics (ROC) curves were obtained as an average of 10 replicates run on MaxEnt, specificity and sensitivity for each parasite was calculated as presented in Fig. 4a,b,c. The average and standard deviation of the Area Under the Curve (AUC) for the 10 replicate runs were 0.992, 0.993, and 0.992 for A. lumbricoides, hookworms and S. stercoralis respectively. These values are indication of excellent performance of the modelling software as an AUC value for greater than 0.80 showed higher sensitivity and specificity for the presence of these parasites. www.nature.com/scientificreports/

Discussion
This study revealed that both climatic and edaphic factors influence the distribution of STHs at varying degree. In this study, mean diurnal temperature range (BIO2), temperature annual range (BIO7), maximum temperature of the warmest month (BIO5), precipitation of the wettest quarter (BIO16)], and soil clay contents were the top five most important factors responsible for their distribution out of the 24 factors investigated. A study carried out in Zimbabwe 35 on the inclusion of edaphic factors in enhancing the distribution of STHs reported that the inclusion of edaphic factors alongside other environmental variables enhance the performance of the predictive models used in determining the distribution of these parasites.
Some studies carried out in Bolivia 34 , China 44 and Nigeria 26 predicted the distribution of these parasites using only environmental variables might have either underestimated or overestimated the distribution. Comparison of this study with a previous study in Nigeria 27 in which only environmental variables were used revealed an improvement in the model i.e. they observed an AUC of 0.948 compared to an AUC of 0.992 in this current study (4.44% increase in model performance). Therefore, inclusion of edaphic variable proved to be important in determining the distribution of STHs especially factors such as soil clay, pH and soil moisture (or soil water retention capacity).
Soil organic carbon was very important in determining the distribution of hookworms. Similar observation was reported in the Zimbabwe 35 , they stated that soil organic carbon is among the highly important factor in determining the distribution of hookworms, this was due to the ecology of hookworms as a parasite that feeds on organic matter. This observation is similar also to a study on Sandy soil in KwaZulu-Natal, South Africa 36 .
For the climatic variables, mean diurnal range (BIO2), precipitation of the wettest quarter (BIO18), precipitation of wettest quarter (BIO16), Temperature annual range (BIO7) and precipitation of wettest month (BIO13) were the factors of most relative importance in the determining the distribution of STHs in Kogi East. An earlier study in Nigeria reported 27 that precipitation of the wettest month was the most important factor in STHs distribution in Nigeria. This result is also in line with previous study 14 on moisture requirement of eggs of parasitic ascarids in mammals. A study in Kebbi State, Nigeria 45 stated the significance of moisture and warm conditions in enhancing the embryonation of the eggs of parasitic helminths.
Mean diurnal temperature range (BIO2), temperature annual range (BIO7), maximum temperature of the warmest month (BIO5), precipitation of the wettest quarter (BIO16)], and soil clay contents were the five factors that exerted most significant influence on the geospatial distribution of STHs in Kogi East, Nigeria.
Sanitation is a crucial environmental factor in eliminating the overall rates of STH infection. However, the higher cost of proper sanitation methods compared to other intervention limit its implementation in many communities particularly where resources are limited 46 . This factor might be responsible for lack of data on sanitation from our study area since they are rural communities with limited resources. This is a limitation of  www.nature.com/scientificreports/ this study. Further research incorporating sanitation parameters to evaluate its effect on the distribution of STHs should be conducted.

Conclusions
Mean diurnal range (BIO2), precipitation of the wettest quarter (BIO16), soil clay content, temperature annual range (BIO7) and maximum temperature of the warmest month (BIO5) are the five most important environmental risk factors that plays significant role in the spatial distribution of STHs in Kogi East, North Central Nigeria. Southern part of Dekina LGA, Eastern part of Ofu LGA, northern part of Igalamela-Odolu LGA, the west-Southern part of Olamaboro LGA and the Eastern part of Omala LGA were areas at high risk of infection with STHs. Policies designed on the monitoring of the distribution of chemotherapy should be reviewed properly by selecting and considering high-risk communities, this will allow maximum achievement of control strategy.

Data availability
The data sets in this study are available from the corresponding author on reasonable request.