MosquitoMap is authored by Desmond Foley, Pollie Rueda and Richard Wilkerson of the Walter Reed Biosystematics Unit (WRBU) based in the Smithsonian Institution, Washington DC. MosquitoMap was designed to be the first dedicated online clearinghouse for global geo-referenced mosquito collection records and species distribution models derived from those records. See a published description of MosquitoMap in the International Journal of Health Geographics. This site is intended to increase our understanding of mosquito species and vector-borne disease distribution. We anticipate that it will also be useful for research on the impact of global warming and anthropogenic changes on emerging infectious diseases and biodiversity. Funding for the creation of MosquitoMap was obtained from the Global Emerging Infections Surveillance and Response System (a Division of the Armed Forces Health Surveillance Center - AFHSC/Div of GEIS Ops) as part of a proposal to conduct ecological niche modelling of mosquito disease vectors. The web application was built by WorldView Solutions Inc. using ArcGIS Server 10 technology. A grant was received from the Global Biodiversity Information Facility (GBIF) early 2008 to assist with the addition of geo-referenced mosquito collection records. These records come from a variety of databases ranging from ready to deploy to those requiring manual entry from written forms. The Mosquitoes of Middle America data formed the nucleus of MosquitoMap, which is intended to be a user-friendly site for all to serve their data, make maps and to query. Orphaned datasets and observation data within the general literature not duplicated in scientific collections are in the process of being gathered, and preserved online.
It is often erroneously assumed that a great deal is known about mosquito systematics and distributions. One reason is that because of their medical importance mosquitoes have been and continue to be thoroughly studied. This can be said perhaps for the 200 or so vector and pest species, about 80 (out of 460) of which are Anopheles malaria vectors. However, there are about 3500 described species (WRBU Mosquito Catalog; Harbach & Howard, 2007, European Mosquito Bulletin, 23: 1-66), most of which are relatively poorly known. Recently, molecular methods have shown, that many Anopheles vector species belong to morphologically similar or indistinguishable species complexes.
The paucity of detailed data on the past and present distribution of mosquito vectors is a major limiting factor for global modeling of vector-borne diseases (e.g. Tatem et al., 2006, Proceedings National Academy of Sciences, USA, 103: 6242-6247). There is also a challenge of choosing appropriate priority taxa that can be used to represent biodiversity in monitoring projects (Pereira & Cooper, 2006,Trends in Ecology and Evolution, 21: 123-129). For this, mosquitoes have been suggested as a priority group within the insects (Raven, 1980, Research Priorities in Tropical Biology, National Academy ofSciences, 116pp).
The web application was built by WorldView Solutions Inc. using ArcGIS Server 9 and was released in 2009. A rebuild, using ArcGIS Server 10 and Microsoft Silverlight became available April 2011. The web application is housed at the University of Kansas Natural History Museum and Biodiversity Research Center.
MosquitoMap is a unique global resource for researchers interested in mosquito-borne diseases and basic questions concerning mosquito distribution and ecology. More specifically, MosquitoMap is an online database of species distribution models and georeferenced species collection events, for individual mosquitoes or pools of mosquitoes of the same species. Collection records and distribution maps come from museum specimens, the literature, and from submissions by other mosquito workers. Data for input to MosquitoMap refers to preserved (vouchered) specimens and human observation data. We rescue mosquito observation data from the literature that does not have associated vouchered specimens so that they will be databased and served. These observations are given temporary Catalog numbers (MMap 1, 2 etc), which comprise part of the datum's global unique identifier. A primary output of MosquitoMap is ‘dots on maps’, or collection records to species for a particular time and place. These collection events may be observations only or may result in vouchered specimens that are stored in museums or other collections. Unlike Museum databases, MosquitoMap is not primarily concerned with data that replicate the record of an individual’s collection, such as occur with exuviae, DNA material, progeny or genitalia preparations. These are considered derivatives and may (or may not) be recorded, unless they are the sole representative of an individual mosquito. Another primary output of MosquitoMap are species distribution models. These may have been developed by a variety of procedures but are usually ecological niche models derived from presence-only data. We provide 'placeholder' models for a variety of vector species to demonstrate the functionality of MosquitoMap. However, these models are not published nor extensively validated, so we invite contributors to send MosquitoMap their published models to allow users a choice of model types.
All records have basic information that accords with the Darwin Core schema including country and subordinate geographic units (using the GADM database), latitude and longitude, spatial accuracy , taxonomic data (including kingdom, family, genus, subgenus and species), an explicit statement of the basis of the record, the collector, collection date, and identification method and date. In all, over 60 data fields are present in MosquitoMap, many with controlled vocabulary terms to assist data searches. Additional data fields include those for larval habitat, and any parasites identified in association with the mosquito (for a complete list of field, see Data Portal). The general description of the methodology for processing and data cleaning the databased georeferenced data is described in Foley et al. (2008. Ecological Entomology). Briefly, databases are divided into those records that have geographic coordinate data and those that do not. Entries are further divided into those of questionable taxonomy, apparent identification failures, and those with unequivocal species identification. Verbatim geographic coordinates are checked to ensure they have the correct sign (+ or -) for their hemisphere of origin, and converted to WGS84 and decimal degrees to 5 decimal places using Geotrans. Specimens with unequivocal identifications and geocodes will be filtered in Microsoft Excel for unique locations, and these point data converted to shape files and simple cleaning routines undertaken by the ‘check coordinates’ option of DIVA-GIS, a “point-in-polygon” method (Chapman, 2005, Principles and Methods of Data Cleaning – Primary Species and Species-Occurrence Data, version 1.0), which identifies points located outside all polygons (i.e. fall in the ocean), and points that do not match relations for the country names (i.e. fall in another country). We also use the GBIF Data Tester tool to further detect records likely to hold erroneous information. Anomalous locations are rechecked and corrected by consulting original collection cards and maps housed at the WRBU or through an online Gazetteer such as the ADL online Gazetteer and the Global Gazetteer. Spatial precision estimates were derived using BioGeoMancer or the Georeferencing calculator. Generic, subgeneric and species names are updated to follow the online Catalog of the Culicidae.