A geographic distribution database of the Neotropical cassava whitefly complex (Hemiptera, Aleyrodidae) and their associated parasitoids and hyperparasitoids (Hymenoptera)

Abstract Whiteflies (Hemiptera, Aleyrodidae) are represented by more than 1,500 herbivorous species around the world. Some of them are notorious pests of cassava (Manihot esculenta), a primary food crop in the tropics. Particularly destructive is a complex of Neotropical cassava whiteflies whose distribution remains restricted to their native range. Despite their importance, neither their distribution, nor that of their associated parasitoids, is well documented. This paper therefore reports observational and specimen-based occurrence records of Neotropical cassava whiteflies and their associated parasitoids and hyperparasitoids. The dataset consists of 1,311 distribution records documented by the International Center for Tropical Agriculture (CIAT) between 1975 and 2012. The specimens are held at CIAT’s Arthropod Reference Collection (CIATARC, Cali, Colombia). Eleven species of whiteflies, 14 species of parasitoids and one species of hyperparasitoids are reported. Approximately 66% of the whitefly records belong to Aleurotrachelus socialis and 16% to Bemisia tuberculata. The parasitoids with most records are Encarsia hispida, Amitus macgowni and Encarsia bellottii for Aleurotrachelus socialis; and Encarsia sophia for Bemisia tuberculata. The complete dataset is available in Darwin Core Archive format via the Global Biodiversity Information Facility (GBIF).


Introduction
Whiteflies (Hemiptera, Aleyrodidae) are represented by more than 1,500 herbivorous species around the world (Hodges and Evans 2005, Evans 2007, 2008. Some of them are notorious pests of cassava (Manihot esculenta), a primary food crop in the tropics (Lebot 2009). Particularly destructive is a complex of Neotropical cassava whiteflies whose distribution remains restricted to their native range (Trujillo et al. 2004, Bellotti et al 2005. Despite their importance, neither their distribution, nor that of their associated parasitoids, is well documented (Evans 2008, Aliaga 2012, da Silva Alonso et al. 2012, Pietrowski et al 2014, Plantwise 2015and Global Biodiversity Information Facility 2015. This paper therefore reports observational and specimen-based occurrence records of Neotropical cassava whiteflies and their associated parasitoids and hyperparasitoids. The dataset consists of 1,311 distribution records documented by the International Center for Tropical Agriculture (CIAT).

Data published through GBIF
http://www.gbif.org/dataset/c6f4c2de-3b71-4ebd-9c98-c21537548f07 Design descriptions: The purpose of this dataset is to broadly and openly share geographic distribution data for the cassava whitefly complex and their associated parasitoids and hyperparasitoids. Prior to this contribution, no records were found of these arthropod species in cassava at the Global Biodiversity Information Facility (2015). To bridge this gap, this paper submits 1,311 distribution records (whiteflies: 841; parasitoids: 466; hyperparasitoids: 4), documented by the International Center for Tropical Agriculture (CIAT) between 1975 and 2012. More than half of these records correspond to specimens preserved at CIAT's Arthropod Reference Collection (CIA-TARC). Most of the whitefly records correspond to Aleurotrachelus socialis Bondar and Bemisia tuberculata Bondar (Fig. 1A). In turn, most parasitoid records belong to Encarsia hispida De Santis, Encarsia sp. and E. sophia (Girault & Dodd) (Fig. 1B). This dataset should be of particular interest to whitefly biologists, cassava entomologists and national plant protection organizations (NPPOs) in tropical countries.

General taxonomic coverage description
Most records were identified to the species level (whiteflies: 97%; parasitoids and hyperparasitoids: 73%) by expert entomologists. Experts identifying more than 20 records were Gregory A. Evans, María del Pilar Hernández, Sueo Nakahara and Louise M. Russell. Whitefly records belong to nine genera and eleven species (Table 1), whereas parasitoid records belong to eight genera and 14 species (Table 1). The dataset also includes four records of the genus Signiphora (Table 1), considered a genus of whitefly hyperparasitoids (Evans 2007

Spatial coverage
General spatial coverage: Most of the distribution records belong to South America (Brazil, Colombia, Ecuador and Venezuela) and Central America (El Salvador, Guatemala, Honduras, Nicaragua and Panama). Colombia and Venezuela are the best represented countries, followed by Brazil and Ecuador (Fig. 1C). There are also seven records of whiteflies from Asia (Lao and Thailand). The distribution maps of principal whiteflies and their parasitoids are shown in Figure 2. Coordinates

Natural collections descriptions
Collection name: CIAT's Arthropod Reference Collection (CIATARC) Specimen preservation method: Specimens are preserved in microslides (whiteflies, parasitoids and hyperparasitoids), tissue beds on dried vials (parasitoids), 70% ethyl alcohol (parasitoids and hyperparasitoids), or in 35 mm plastic slide mounts (whiteflies). These samples are deposited within cabinet drawers maintained at 21.0 ± 0.4 °C and 47.6 ± 8.6% relative humidity. They are sorted numerically by species and country of origin.
Curatorial unit: 1601 with an uncertainty of 0.

Method step description:
The dataset integrates two data flows: observational records and specimen-based records, identified either to genus or to species. The former were digitized from field diagnostic forms recorded by personnel extensively trained in identification of whiteflies and parasitoids identification. These identifications, however, were likely conducted on site without mounting and preserving samples. Alternatively, these observations may correspond to properly-mounted but lost specimens.
In either case, we are significantly confident on these identifications due to relatively clear macroscopic differences in our focal taxa (Caballero 1994, Fernández andSharkey 2006). Still, conservative users of our database may prefer to rely only on genus-level identifications of these records. On the other hand, the specimen-based records belong to verifiable samples properly-preserved at CIATARC. Guidelines of Martin (1987) and Hodges and Evans (2005) were followed for whitefly slide preparations, and Noyes (1982) for parasitoid and hyperparasitoid preparations. Unique accession numbers were assigned to all records. All biodiversity data available (i.e. specimen, species identification, name of determiner, sex, locality, date, habitat, host, collector and observations) were digitized in a Microsoft Excel 2010 spreadsheet adopting the Darwin Core Archive format v1.2 (Wieczorek et al. 2012). We updated locality fields (e.g., district, municipality) using the most current names and classifications of administrative divisions used by each country (e.g. http://www.dane.gov.co/Divipola/ for Colombia, http://www.inec.gob. ec/estadisticas/?option=com_content&view=article&id=80 for Ecuador, etc. [accessed 14 November 2014]). Based on their locality names, we then geocoded the records using Google Maps (https://maps.google.com/), Geolocate (http://www.museum.tulane. edu/geolocate/), GeoNames (http://www.geonames.org/) or with georeference indicated in scientific articles (Calderón et al. 1994, Eiszner et al. 1996, Navia Estrada et al. 2006, Cuadros et al. 2011, Gutiérrez R. et al. 2011. GPS coordinates were converted to decimal degrees. The dataset with metadata was uploaded to the Integrated Publishing Toolkit (IPT) of the Colombia node of Global Biodiversity Information Facility (GBIF) (http://www.gbif.org/dataset/c6f4c2de-3b71-4ebd-9c98-c21537548f07).
Sampling description: The records in the dataset have been documented in three ways: 1) Records from CIAT's initial field explorations to document pests in cassava (CIAT 1974(CIAT , 19850.7% records, between 1975-1989. 2) Records documented during the "Biological Control of Whiteflies by Indigenous Natural Enemies for Major Food Crops in the Neotropics Projects" and participation in "Global Whitefly IPM Project" led by CIAT, Instituto Nacional de Investigaciones Agropecuarias (INIAP), Centro Nacional de Investigaciones Agropecuarias (CENIAP), Empresa Brasileira de Pesquisa Agropecuária (EMBRAPA), The University of Florida and Corporación Colombiana de Investigacion Agropecuaria (Corpoica) (CIAT 1995, Bellotti et al. 1996, 1999, 2000, Bellotti 2001, Arias and Bellotti 2002, CIAT et al. 1998, Castillo 1996, López-Ávila et al. 2001, Hernandez and Bellotti 2002, Holguín et al. 2002, Hernández et al. 2004, 2009, Trujillo et al. 2004, Herrera et al. 200695.7% records, between 1990. 3 Records from other sources; including field inspections and collections conducted during routine farm visits by CIAT personnel, and specimens submitted to CIATARC by fellow institutions and researchers (Adriano Muñoz and Gerardino Perez, pers. comm. November 29, 2014;2.6% records between 2008-2012. The records resulted from one of two sampling methods. The first method was designed to identify parasitoids associated with dominant whitefly species on farmers' fields. One middle-canopy leaf infested with whiteflies was collected from each of 40-100 randomly-selected plants per field. A disc of 2.54 cm 2 was excised from the leaf lobe with the highest density of whitefly pupae. The single most abundant whitefly species per disc was identified and individuals not belonging to that species were eliminated by puncturing them with a needle. The disc samples were then individually placed in 25-ml glass vials and held for 2-3 days at 24.5 ± 4 °C and 70 ± 5% relative humidity under laboratory conditions until parasitoids emerged (Bellotti et al. 1999, 2000, Trujillo et al. 2004). The second method corresponds to opportunistic collections during routine farm visits by CIAT personnel, when leaves infested with whitefly pupae would be collected in vials with 70% alcohol and submitted to the CIATARC for subsequent identification (Herrera et al. 2006). All formally-submitted samples were mounted and are preserved at the CIATARC. The database does not indicate which sampling method was used for each record.
Quality control description: Record validation and cleaning was incorporated at several steps of the documentation process, following guideless by Chapman (2005a, b). The scientific names on labels were checked with a taxonomic thesaurus developed by Aymer Andrés Vásquez-Ordóñez, Juan David Escobar-Prieto and Dario Paz-Jojoa. This thesaurus compiled all known synonyms and spelling variants of the scientific names used for our focal species. Scientific names were assigned in accordance to current taxonomic trends (whiteflies: Evans 2008; parasitoids and hyperparasitoids: Woolley 1988, Polaszek et al. 2004, Evans 2007, Johnson 2007, Noyes 2014associated plants: Tropicos 2014). Geographic coordinates were verified using the "Check Coordinates" function in DIVA-GIS (Hitmans et al. 2001). For this last step, we relied on the Global Administrative Unit Layers (GAUL) shape file developed by the Food and Agriculture Organization of the United Nations (FAO 2015), and official shape of administrative division of Brazil, Colombia, Ecuador and Venezuela (IBGE 2007, INEC 2011, SIGOT 2011, IVIC 2007.

Dataset description
Object name: Darwin Core Archive cassava whiteflies complex and their associated parasitoids and hyperparasitoids: data of the CIAT's Arthropod Reference Collection of International Center for Tropical Agriculture (CIAT).