Applications of ecological niche modeling to enhance medical threat assessment and disease control and prevention strategies.
Preventive health services
|Publication:||Name: U.S. Army Medical Department Journal Publisher: U.S. Army Medical Department Center & School Audience: Professional Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2009 U.S. Army Medical Department Center & School ISSN: 1524-0436|
|Issue:||Date: July-Sept, 2009|
|Product:||Product Code: 8000140 Health Problems Prevention; 9105230 Health Problems Prevention Programs NAICS Code: 621999 All Other Miscellaneous Ambulatory Health Care Services; 92312 Administration of Public Health Programs|
One of the major challenges for military preventive medicine is in
developing medical threat assessments both prior to and during
deployment. When it comes to vector-borne diseases, this can be
especially problematic. In many areas, it is either impractical or
impossible to conduct a thorough predeployment surveillance program.
This leads to reliance on products, such as the Disease Vector Ecology
Profiles, that may be out of date or too general in nature.
In recent years, the use of remotely sensed data (ie, satellite imagery) has gained popularity in many areas of biology, including epidemiology. (1) Along with this, a number of modeling approaches have been developed to answer basic questions about the distribution, ecology, and potential range of a species in the absence of comprehensive sampling data. These techniques are commonly referred to as ecological niche modeling or species distribution modeling. Niche modeling holds the potential to be a powerful tool for the military preventive medicine community in developing more robust risk assessments for vector-borne disease and better targeting surveillance, control, and education efforts.
Joseph Grinell originally presented the idea that species inhabit specific areas (the niche), defined by the species' biological requirements for survival. (2) The ecological niche is determined by 4 interacting factors. (3-5) These are:
* Biotic factors: Interactions with other species including: predators, pathogens, prey, parasites, and vegetation.
* Abiotic factors: The environmental factors that determine whether or not a species can survive in a given location (climate, soil, etc.).
* The ability of a species to disperse to new areas. This includes any physical barriers to dispersion (such as mountain ranges, bodies of water, and deserts) as well as the biological capacity or propensity of a species to disperse.
* The ability of the species to adapt to new environments.
Together, these factors determine the actual and potential geographic distribution of a species. (3) Niche modeling techniques aim to answer a variety of questions based on the knowledge of these factors for a given species.
Figure 1 presents a diagrammatic representation of the factors that determine the niche for a given organism. The area in which all 3 circles overlap is the actual niche of a species. The areas where only 2 circles overlap are the potential niche. For instance, the area labeled "D" in this example is currently uninhabitable by a species due to abiotic factors (such as temperature or precipitation). If the abiotic factors limiting the range of the species are changed (such as through an El Nino or La Nina event, or some other form of climate change) so that they are conducive to that species, the niche could shift to include that area.
Niche modeling has been increasingly applied to disease ecology. (5) Models of monthly predictions of Dengue fever in Mexico have been created based on mosquito activity. (6) Niche modeling has been used to help prevent anthrax in wildlife and livestock, by predicting the expected distribution of the pathogen, Bacillus anthracis (Cohn) in the environment. (7) Niche models of malaria vectors in the Anopheles gambiae complex have been developed for undersampled regions of Africa. (8) The distributions of Triatoma spp vectors of Chagas' disease in the Americas have been examined in an effort to better refine vector control programs. (4,9) A number of studies have focused on both present and potential distributions of Lutzomyia spp vectors of cutaneous and visceral leishmaniasis in Central and South America. (10-13) Niche modeling allows the user to develop maps showing predicted distribution of an organism based on current and projected data.
[FIGURE 1 OMITTED]
One of the reasons that niche modeling has grown in use is the ability to use records from a variety of sources, including museum records, literature reports, and direct sampling. Many of the computer algorithms that have been developed for this modeling can create distribution models without actual absence records. This ability to use only presence records is what enables the creation of models for areas with poor or incomplete sampling. A number of methods have been developed to perform niche modeling. However, the most prevalent are GARP (genetic algorithm rule-set predictor) and MAXENT (maximum entropy) techniques. (14) Software for both applications is available freely on the internet and is simple to install and use. However, knowledge of geographic information system software is helpful in developing and refining the risk maps created by these techniques.
Using Niche Modeling to Determine Sand Fly Distribution
Niche modeling has been used to develop a map of the predicted distribution of Phlebotomus papatasi (Scopoli) which is a vector of sand fly fever virus and Leishmania major, a protozoan that causes cutaneous leishmaniasis (M.C-M., unpublished data, 2009). Both sand fly fever and cutaneous leishmaniasis are important disease threats to military operations in the middle east and Mediterranean regions. (15-17) The abiotic factors affecting this species include temperature, precipitation, and elevation. The biotic factors relevant to this species include availability of acceptable blood meal sources, level of predation, and the availability of the appropriate vegetation to provide a sugar source. The limits to dispersal include mountain ranges, bodies of water, and the distance the insect can fly. Using Figure 1 as a reference, the area labeled "A" represents the actual distribution of this species. If military forces were entering an area that did not meet the requirements of area A, the diseases vectored by this species would not be a concern. However, if military forces were operating in this area, preventive medicine personnel could plan appropriately for prevention and control of the vectors, and Soldier training could be adjusted to address the actual threat.
Figure 2 shows the predicted distribution of P papatasi across a portion of the Middle East. This map was developed using the niche modeling software Maxent, version 3.2.1. It incorporates sand fly collection records (both from direct collections and records in the literature), temperature, precipitation, and land use data (M.C-M., unpublished data, 2009). While the map is similar to the one in the Disease and Vector Ecology Profile for cutaneous leishmaniasis in the same area (Figure 3), it provides more refined information about the distribution of a specific vector of concern and is easily updated. Such models could be developed rapidly, targeting specific areas of interest to provide the most realistic picture of the disease risk to deployed or deploying military personnel.
[FIGURE 2 OMITTED]
Using Niche Modeling to Determine Mosquito Distribution
Researchers at the Walter Reed Biosystematics Unit are developing a novel online tool called MosquitoMap (19) that provides a clearinghouse of information on mosquito distribution and risk of mosquito-borne diseases worldwide. The application incorporates niche models developed by individuals across the modeling community. Users may use the tool to determine what mosquito species are present, overlay distributional models of specific mosquito species, and/or overlay models of mosquito-borne diseases. Disease models currently incorporated in MosquitoMap are Plasmodium vivax and Plasmodium falciparum malaria, yellow fever, dengue fever, Japanese encephalitis, and Rift Valley fever. A useful component of this effort is the development of the "Mal-area calculator" which determines the areas where the distribution of human populations, Plasmodium spp parasites, and disease vectors overlap in order to better gauge the real malaria risk in a given area. (19) Figure 4 is a representation of the MosquitoMap model for the Republic of Korea, displaying the distribution of Anopheles sinensis (Wiedemann), a vector of P vivax malaria.
[FIGURE 3 OMITTED]
Once complete, this tool will allow public health personnel to better determine the threat of mosquito-borne disease in an area of interest. This sort of global or regional plan could be applied to other vectors and vector-borne diseases. Such a resource could provide preventive medicine personnel targeted maps to use in planning disease control and prevention strategies. Ultimately, this sort of comprehensive effort could help to target use of prophylaxis, allow military entomologists to refine surveillance and control measures, and help emphasize the risk and importance of disease threats to an operation.
Niche modeling tools have potential to help preventive medicine personnel better understand the medical threat to our forces in areas to which they are deployed or may be deploying. Some of the potential applications have already been highlighted.
These tools may be used to assist in the development of the predeployment medical threat assessment. For some operations (such as the movement of combat forces), it may be impossible or impractical to conduct a thorough, on the ground, predeployment site survey of areas in which US military units will be engaged. In addition, many areas in which we operate have incomplete or inaccurate data regarding the threat of disease. In these instances, niche modeling could help fill in the information gaps by modeling the relative risk of key vector species or vector-borne diseases across an area of interest.
Niche modeling can be used to help identify the effects of habitat or climate change on a species or disease. This application would be beneficial in projecting disease threat in areas where the environment is changing (ie, as a result of a hurricane or tsunami). Modeling with the goal of predicting future populations of vector species could assist public health planners in addressing ongoing preventive medicine concerns over weeks or months after such a habitat alteration scenario.
[FIGURE 4 OMITTED]
At a finer scale, these techniques could help target vector surveillance efforts, determine the best layout of a base in order to minimize contact with a vector, or identify appropriate engineering controls for a target species. If these tools are available to the military entomologists on the ground, vector control strategies could be enhanced through the use of remote imagery and predictive models to increase the overall efficacy of surveillance and control strategies.
A great deal of work remains for the military to employ these tools to their full potential. There should be some effort to coordinate public health research utilizing remote sensing and predictive modeling across the military in order to most effectively and efficiently exploit these technologies. Work should be targeted and focused on priority vectors, diseases, and regions in order to create a comprehensive and useful tool for the military public health community. With some forethought and deliberate planning, these tools hold tremendous potential to improve our ability to identify the risk to our forces and develop better disease prevention and control strategies during all phases of deployment.
(1.) Peterson AT. Ecological niche modeling and spatial patterns of disease transmission. Emerg Infect Dis. 2006;12(12):1822-1826.
(2.) Grinnell J. Field tests of theories concerning distributional control. Am Nat. 1917;51:115-128.
(3.) Soberon J, Peterson AT. Interpretation of models of fundamental ecological niches and species' distributional areas. Biodiversity Informatics. 2005;2:1-10.
(4.) Costa J, Peterson AT, Beard CB. Ecological niche modeling and differentiation of populations of Triatoma brasiliensis Neiva, 1911, the most important Chagas' disease vector in northeastern Brazil (Hemiptera, Reduviidae, Triatominae). Am J Trop Med Hyg. 2002;67(5):516-520.
(5.) Peterson A. Biogeography of diseases: a framework for analysis. Naturwissenschaften. 2008;95:483-491.
(6.) Peterson AT, Martinez-Campos C, Nakazawa Y, Martinez-Meyer E. Time-specific ecological niche modeling predicts spatial dynamics of vector insects and human dengue cases. Trans R Soc Trop Med Hyg. 2005;99:647-655.
(7.) Blackburn J, McNyset K, Curtis A, Hugh-Jones M. Modeling the geographic distribution of Bacillus anthracis, the causative agent of anthrax disease, for the contiguous United States using predictive ecological niche modeling. Am J Trop Med Hyg. 2007;77(6):1103-1110.
(8.) Levine RS, Peterson AT, Benedict MQ. Geographic and ecological distributions of the Anopheles gambiae complex predicted using a genetic algorithm. Am J Trop Med Hyg. 2004;70(2):105-109.
(9.) Peterson AT, Sanchez-Cordero V, Beard CB, Ramsey JM. Eclogical niche modeling and potential reservoirs for chagas disease, Mexico. Emerg Infect Dis. 2002;8 (7):662-667.
(10.) Peterson AT, Shaw J. Lutzomyia vectors for cutaneous leishmaniasis in Southern Brazil: ecological niche models, predicted geographic distributions, and climate change effects. Int J Parasitol. 2003;33:919-931.
(11.) Peterson AT, Pereira RS, Neves V. Using epidemiological survey data to infer geographic distributions of leishmaniasis vector species. Rev Soc Bras Med Trop. 2004;37:10-14.
(12.) da Costa SM, Cechinel M, Bandiera V, Zannuncio JC, Lainson R, Rangel EF. Lutzomyia (Nyssomyia) whitmani s.l. (antunes & Coutinho, 1939) (Diptera: Psychodidae: Phlebotominae): geographical distribution and the epidemiology of American cutaneous leishmaniasis in Brazil--mini review. Mem Inst Oswaldo Cruz. 2007;102(2):149-153.
(13.) Nieto P, Malone JB, Bavia ME. Ecological niche modeling for visceral leishmaniasis in the state of Bahia, Brazil, using genetic algorithm for rule-set prediction and growing degree day-water budget analysis. Geospat Health. 2006;1:115-126.
(14.) Elith J, Graham C, Anderson R, et al. Novel methods improve prediction of species' distributions from occurrence data. Ecography. 2006;29:129-151.
(15.) Coleman RE, Burkett DA, Putnam JL, et al. Impact of Phlebotomine sand flies on US military operations at Tallil Air Base, Iraq: 1. Background, Military Situation, and Development of a "Leishmaniasis Control Prorgam". J Med Entomol. 2006;43(4):647-662.
(16.) Aronson N, Sanders J, Moran K. In harm's way: infections in deployed American military forces. Clin Infect Dis. 2006;43:1045-1051.
(17.) Ellis SB, Appenzeller G, Lee H, et al. Outbreak of sandfly fever in Central Iraq, September 2007. Mil Med. 2008;173(10):949-953.
(18.) Defense Pest Management Information Analysis Center. Regional Disease Vector Ecology Profile: The Middle East. Silver Spring, MD: Armed Forces Pest Management Board; 1999. Available at: http:// www.afpmb.org/pubs/dveps/mid_east.pdf. Accessed August 6, 2009.
(19.) Foley D, Wilkerson R. MosquitoMap Website. Available at: http://www.mosquitomap.org/ index.htm. Accessed February 27, 2009.
MAJ Colacicco-Mayhugh is Chief, Department of Sand Fly Biology, Division of Entomology, Walter Reed Army Institute of Research, Silver Spring, Maryland.
MAJ Michelle Colacicco-Mayhugh, MS, USA
|Gale Copyright:||Copyright 2009 Gale, Cengage Learning. All rights reserved.|