Ecological niche modeling-based prediction on transmission risk of visceral leishmaniasis in the extension region of Loess Plateau, China

doi:10.12140/j.issn.1000-7423.2021.02.015

Abstract

Abstract:

Objective The ecological niche ensemble model was used to analyze and predict the distribution of transmission risk area of visceral leishmaniasis in Shanxi and Henan Provinces in the extension region of the Loess Plateau. Methods The locations with reported cases of visceral leishmaniasis in Shanxi and Henan Provinces from 2015 to 2019 were selected as the distribution sites, from which nine niche models were constructed based on 12 environmental variables in 3 categories and the distribution structure of ecological areas, including generalized linear models (GLM), generalized additive models (GAM), multivariate adaptive regression splines (MARS), generalized boosted models (GBM), classification tree analysis (CTA), flexible discriminant analysis (FDA), artificial neural networks (ANN), Random Forest (RF), and maximum entropy (MaxEnt). The ensemble model was established based on the area under the curve (AUC) of Receiver Operating Characteristic and true skill statistic (TSS). The transmission risk of visceral leishmaniasis was predicted in Shanxi and Henan Provinces. Results The 9 models had statistically significant difference in performance (AUC value, H = 35.742, P < 0.05; TSS value, H = 23.620, P < 0.05), among them, the RF model (AUC = 0.950, TSS = 0.829) and GBM models (AUC = 0.943, TSS = 0.803) performed better than the other single model. The performance of the ensemble model was better than the single model. The transmission risk of visceral leishmaniasis was predicted to be distributed in the Yanshan-Taihang Mountain Deciduous Broad-leaved Forest Ecological Area and the Fen-Wei Basin Agricultural Ecological Area. The risk areas of Shanxi Province accounted for 30.30% of the province’s area, and could be categorized into low-risk (12.99%), medium-risk (13.93%), and high-risk areas (3.37%). The high-risk areas of Shanxi Province were mainly located in the central and southern part of Yangquan City, North of Changzhi City, and South of Linfen City. The risk areas in Henan Province accounted for 4.68% of the Province’s area, and can be divided into the low-risk (3.51%), medium-risk (0.94%), and high-risk areas (0.23%). The high-risk areas of Henan Province were mainly located in the West of Anyang City. Conclusion The transmission risk of visceral leishmaniasis in Shanxi and Henan provinces in the extension region of the Loess Plateau shows a overall scattering and local clustering status in recent years. The ensemble ecological niche model has the potential in analysis and prediction of the disease, being able to provide scientific basis for prevention and control in key areas of leishmaniasis.

Key words: Visceral leishmaniasis, Extension regions of Loess Plateau, Ecological niche modeling, Transmission risk

CLC Number:

R531.6

GONG Yan-feng, HU Xiao-kang, ZHOU Zheng-bin, ZHU Hui-hui, HAO Yu-wan, WANG Qiang, ZHANG Yi, LI Shi-zhu. Ecological niche modeling-based prediction on transmission risk of visceral leishmaniasis in the extension region of Loess Plateau, China[J]. CHINESE JOURNAL OF PARASITOLOGY AND PARASITIC DISEASES, 2021, 39(2): 218-225.

Figures/Tables 5

Table 1

Fig. 1

Table 2

Table 3

Fig. 2

References 34

[1]	Alvar J, Yactayo S, Bern C. Leishmaniasis and poverty[J]. Trends Parasitol, 2006,22(12):552-557. doi: 10.1016/j.pt.2006.09.004
[2]	Oryan A, Akbari M. Worldwide risk factors in leishmaniasis[J]. Asian Pac J Trop Med, 2016,9(10):925-932. doi: S1995-7645(16)30157-2 pmid: 27794384
[3]	Lun ZR, Wu MS, Chen YF, et al. Visceral leishmaniasis in China: an endemic disease under control[J]. Clin Microbiol Rev, 2015,28(4):987-1004. doi: 10.1128/CMR.00080-14
[4]	Wang JY, Cui G, Chen HT, et al. Current epidemiological profile and features of visceral leishmaniasis in People’s republic of China[J]. Parasit Vectors, 2012,5:31. doi: 10.1186/1756-3305-5-31
[5]	Zheng CJ, Xue CZ, Wu WP, et al. Epidemiological characteristics of Kala-azar disease in China, during 2005—2015[J]. Chin J Epidemiol, 2017,38(4):431-434. (in Chinese) doi: 10.3760/cma.j.issn.0254-6450.2017.04.004 pmid: 28468057
	( 郑灿军, 薛垂召, 伍卫平, 等. 我国2005—2015年黑热病报告病例流行特征分析[J]. 中华流行病学杂志, 2017,38(4):431-434.) pmid: 28468057
[6]	Chen HM, Chen HY, Gao JP, et al. Ecological niches of sandfly (Diptera ∶ Psychodidae) in the extension region of Loess Plateau, China: an endemic focus of visceral leishmaniasis[J]. Chin J Vector Biol Control, 2019,30(6):597-602. (in Chinese)
	( 陈翰明, 陈辉莹, 高景鹏, 等. 我国黄土高原延伸地带利什曼病流行区的传播媒介小生境调查[J]. 中国媒介生物学及控制杂志, 2019,30(6):597-602.)
[7]	Zhou ZB, Li YY, Zhang Y, et al. Prevalence of visceral leishmaniasis in China in 2019[J]. Chin J Parasitol Parasit Dis, 2020,38(5):602-607. (in Chinese)
	( 周正斌, 李元元, 张仪, 等. 2019年我国内脏利什曼病疫情分析[J]. 中国寄生虫学与寄生虫病杂志, 2020,38(5):602-607.)
[8]	Zhou ZB, Li YY, Zhang Y, et al. Prevalence of visceral leishmaniasis in China in 2018[J]. Chin J Parasitol Parasit Dis, 2020,38(2):175-180, 187. (in Chinese)
	( 周正斌, 李元元, 张仪, 等. 2018年全国内脏利什曼病疫情分析[J]. 中国寄生虫学与寄生虫病杂志, 2020,38(2):175-180, 187.)
[9]	Falcão de Oliveira E, Galati EAB, Oliveira AGD, et al . Ecological niche modelling and predicted geographic distribution of Lutzomyia cruzi, vector of Leishmania infantum in South America[J]. PLoS Negl Trop Dis, 2018,12(7):e0006684. doi: 10.1371/journal.pntd.0006684
[10]	Chaves LF, Calzada JE, Valderrama A, et al. Cutaneous leishmaniasis and sand fly fluctuations are associated with el niño in panamá[J]. PLoS Negl Trop Dis, 2014,8(10):e3210. doi: 10.1371/journal.pntd.0003210
[11]	Ready PD. Leishmaniasis emergence and climate change[J]. Rev Sci Tech, 2008,27(2):399-412. doi: 10.20506/rst.issue.27.2.38
[12]	Lin XL, Xiao H, Tian HY. Application of niche model in risk forecast of infectious diseases[J]. Chin J Prev Med, 2013,47(4):294-296. (in Chinese)
	( 林晓玲, 肖洪, 田怀玉. 生态位模型在传染病风险预测中的应用[J]. 中华预防医学杂志, 2013,47(4):294-296.)
[13]	de Santana Martins Rodgers M, Bavia ME, Fonseca EOL, et al. Ecological niche models for sand fly species and predicted distribution of Lutzomyia longipalpis (Diptera ∶ Psychodidae) and visceral leishmaniasis in Bahia state, Brazil[J]. Environ Monit Assess, 2019,191(Suppl 2):331. doi: 10.1007/s10661-019-7431-2 pmid: 31254126
[14]	Abdullah AYM, Dewan A, Shogib MRI, et al. Environmental factors associated with the distribution of visceral leishmaniasis in endemic areas of Bangladesh: modeling the ecological niche[J]. Trop Med Heal, 2017,45(1):1-15.
[15]	Artun O. Ecological niche modeling for the prediction of cutaneous leishmaniasis epidemiology in current and projected future in Adana, Turkey[J]. J Vector Borne Dis, 2019,56(2):127-133. doi: 10.4103/0972-9062.263726
[16]	Thuiller W, Lafourcade B, Engler R, et al. BIOMOD: a platform for ensemble forecasting of species distributions[J]. Ecography, 2009,32(3):369-373. doi: 10.1111/eco.2009.32.issue-3
[17]	Leta S, Fetene E, Mulatu T, et al. Modeling the global distribution of Culicoides imicola: an ensemble approach[J]. Sci Rep, 2019,9(1):14187. doi: 10.1038/s41598-019-50765-1
[18]	Heikkinen RK, Marmion M, Luoto M. Does the interpolation accuracy of species distribution models come at the expense of transferability?[J]. Ecography, 2012,35(3):276-288. doi: 10.1111/ecog.2012.35.issue-3
[19]	Allouche O, Tsoar A, Kadmon R. Assessing the accuracy of species distribution models: prevalence, kappa and the true skill statistic (TSS)[J]. J Appl Ecol, 2006,43(6):1223-1232. doi: 10.1111/jpe.2006.43.issue-6
[20]	Mędrzycki P, Jarzyna I, Obidziński A, et al. Simple yet effective: historical proximity variables improve the species distribution models for invasive giant hogweed (Heracleum mantegazzianum s.l.) in Poland[J]. PLoS One, 2017,12(9):e0184677. doi: 10.1371/journal.pone.0184677
[21]	Hu XK, Hao YW, Xia S, et al. Detection of schistosomiasis transmission risks in Yunnan Province based on ecological niche modeling[J]. Chin J Parasitol Parasit Dis, 2020,38(1):80-86, 94. (in Chinese)
	( 胡小康, 郝瑜婉, 夏尚, 等. 基于生态位模型的云南省血吸虫病传播风险探测研究[J]. 中国寄生虫学与寄生虫病杂志, 2020,38(1):80-86, 94.)
[22]	Yang SL, Song LJ, Ma GQ, et al. Investigation of 4 cases of kala-azar in Anyang City, Henan Province[J]. Chin J Endem, 2018,37(12):1027. (in Chinese)
	( 杨书丽, 宋录军, 马改青, 等. 河南省安阳市4例黑热病病例调查[J]. 中华地方病学杂志, 2018,37(12):1027.)
[23]	Elith J, Graham C, Anderson R, et al. Novel methods improve prediction of species’ distributions from occurrence data[J]. Ecography, 2006,29(2):129-151. doi: 10.1111/j.2006.0906-7590.04596.x
[24]	Qiao HJ, Soberón J, Peterson AT. No silver bullets in correlative ecological niche modelling: insights from testing among many potential algorithms for niche estimation[J]. Methods Ecol Evol, 2015,6(10):1126-1136. doi: 10.1111/mee3.2015.6.issue-10
[25]	Townsend J. Mapping disease transmission risk: enriching models using biogeography and ecology[J]. Emerg Infect Dis, 2015,21(8):1489. doi: 10.3201/eid2108.150665
[26]	Dai PF, Tian XD, Zhao JY, et al. The effectiveness comparative study of three different devices in trapping sandflies in different areas of Shanxi Province, China[J]. Chin J Vector Biol Control, 2020,31(2):212-214, 218. (in Chinese)
	( 代培芳, 田晓东, 赵俊英, 等. 3种不同器械在山西省不同地区诱捕白蛉效果比较研究[J]. 中国媒介生物学及控制杂志, 2020,31(2):212-214, 218.)
[27]	Yang JK, Li F, Liu LL, et al. Effect of sand-fly control on interruption of leishmaniasis transmission[J]. Chin Trop Med, 2014,14(8):932-934. (in Chinese)
	( 杨俊克, 李凡, 刘林林, 等. 防止白蛉叮吸以降低犬源型黑热病传播风险的效果评价[J]. 中国热带医学, 2014,14(8):932-934.)
[28]	Li YF, Zhong WX, Zhao GH, et al. Prevalence and control of kala-azar in China[J]. J Pathog Biol, 2011,6(8):629-631. (in Chinese)
	( 李玉凤, 仲维霞, 赵桂华, 等. 我国黑热病的流行概况和防治现状[J]. 中国病原生物学杂志, 2011,6(8):629-631.)
[29]	Han S, Wu WP, Xue CZ, et al. Endemic status of visceral leishmaniasis in China from 2004 to 2016[J]. Chin J Parasitol Parasit Dis, 2019,37(2):189-195. (in Chinese)
	( 韩帅, 伍卫平, 薛垂召, 等. 2004—2016年中国内脏利什曼病疫情分析[J]. 中国寄生虫学与寄生虫病杂志, 2019,37(2):189-195.)
[30]	Cross ER, Hyams KC. The potential effect of global warming on the geographic and seasonal distribution of Phlebotomus papatasi in southwest Asia[J]. Environ Health Perspect, 1996,104(7):724-727. doi: 10.1289/ehp.96104724
[31]	Zhang P, Shen ZY, Zhang YP, et al. Research progress of clinical epidemiology, prevention and treatment on visceral leishmaniasis[J]. Med J NDFNC, 2019,40(11):703-708. (in Chinese)
	( 张鹏, 沈兆媛, 张亚萍, 等. 我国内脏利什曼病临床流行病学特征与防治研究现状[J]. 西北国防医学杂志, 2019,40(11):703-708.)
[32]	Zheng CJ, Wang LY, Xu X, et al. Visceral leishmaniasis in China during 2004—2007[J]. Chin J Parasitol Parasit Dis, 2009,27(4):344-346. (in Chinese)
	( 郑灿军, 王立英, 许翔, 等. 2004—2007年我国内脏利什曼病流行情况[J]. 中国寄生虫学与寄生虫病杂志, 2009,27(4):344-346.)
[33]	Li H, Zhao YN, Zhao LQ, et al. A case of recurrent Kala-azar in Shanxi Province[J]. Chin J Endem, 2019,38(3):247-248. (in Chinese)
	( 李红, 赵亚楠, 赵丽琴, 等. 山西省1例再发内脏利什曼病患者诊治体会[J]. 中华地方病学杂志, 2019,38(3):247-248.)
[34]	Zheng CJ, Fu JY, Li Z, et al. Spatiotemporal variation and hot spot detection of visceral leishmaniasis disease in Kashi prefecture, China[J]. Int J Environ Res Public Health, 2018,15(12):E2784.

变量分类 Variable classification	变量名 Variable name	含义 Definition	分辨率 Resolution	年份 Year
气候因素 Climatic factor	AAT0	≥ 0 ℃年积温 ≥ 0 ℃ Annual accumulated temperature	500 m	1950—1990
	AAT10	≥ 10 ℃年积温 ≥ 10 ℃ Annual accumulated temperature	500 m	1950—1990
	AR	干燥度Aridity	500 m	1950—1990
	IM	湿润指数Index of moisture	500 m	1950—1990
	AAP	年平均降水量 Average annual precipitation	1 km	2004—2015
	AAT	年平均气温 Average annual temperature	1 km	2004—2015
地理因素 Geographical factor	LF	地貌类型Landform	30 m	2010
	EL	海拔高度Elevation	1 km	2000
	LU	土地利用类型Landuse	1 km	2015
	ANDVI	年度植被指数Annual vegetation index	1 km	2004—2015
社会经济因素 Social and economic factor	GDP	国内生产总值Gross domestic product	1 km	2015
社会经济因素 Social and economic factor	DP	人口密度Density of population	1 km	2015

模型Model	AUC值AUC value	TSS值TSS value
ANN	0.871 ± 0.057	0.677 ± 0.057
CTA	0.860 ± 0.059	0.685 ± 0.059
FDA	0.883 ± 0.055	0.728 ± 0.055
GAM	0.795 ± 0.088	0.601 ± 0.088
GBM	0.943 ± 0.035	0.803 ± 0.035
GLM	0.873 ± 0.098	0.728 ± 0.098
MARS	0.849 ± 0.085	0.648 ± 0.085
MaxEnt	0.813 ± 0.056	0.625 ± 0.056
RF	0.950 ± 0.027	0.829 ± 0.027

变量 Variable	环境变量重要性百分比/% Environmental variable importance percentage/%
变量 Variable	GLM	GBM	GAM	CTA	ANN	MARS	FDA	RF	MaxEnt
AAP	8.1	14.5	10.3	32.4	13.6	13.0	11.8	13.5	2.7
IM	18.4	9.6	9.2	25.1	8.2	11.3	4.1	13.1	1.5
GDP	0.8	4.8	2.6	2.0	6.9	1.2	0	10.3	15.3
EL	17.6	28.7	10.6	23.6	7.7	7.4	13.4	10.2	12.0
AAT	15.1	12.4	16.8	13.0	3.6	26.8	36.0	9.8	23.6
AAT0	10.1	1.5	14.2	0	25.0	13.6	13.2	9.2	0.7
ANDVI	3.3	9.0	4.6	0	0	2.0	3.5	8.5	9.4
AR	9.7	6.3	9.6	0	4.1	9.0	2.0	8.5	1.4
DP	0.7	9.1	2.8	0	4.5	3.8	0	7.8	10.5
AAT10	8.7	3.2	10.9	2.3	23.2	6.4	10.2	6.7	2.0
LF	4.7	1.0	4.2	1.6	2.4	1.6	2.2	2.3	16.4
LU	2.9	0	4.3	0	0.9	3.7	3.6	0.1	4.5