A location-specific spreadsheet for estimating Zika risk and timing for Zika vector surveillance, using U.S. military facilities as an example

Local Zika virus transmission in the United States involving one or both of the known vector species, Aedes aegypti and Ae. albopictus, is of major concern. To assist efforts to anticipate the risks of transmission, we developed an Excel spreadsheet tool that uses vector and virus temperature thresholds, remotely sensed maximum temperature, and habitat suitability from models to answer the questions: “is Zika transmission likely here?” and “when should we conduct vector surveillance?”. An example spreadsheet, updated regularly and freely available, uses near real-time and forecast temperature data to generate guidance, based on a novel four level Zika risk code, for 733 U.S. military facilities in the 50 states, the District of Columbia, and the territories of Guam and Puerto Rico.


INTRODUCTION 24
In 2016, Zika virus disease and congenital infections became nationally notifiable 25 conditions in the United States (Council of State and Territorial Epidemiologists, 2016). 26 A total of 2,382 confirmed and probable cases of ZIKAV disease with illness onset were 27 reported to ArboNET, the U.S. national arboviral surveillance system managed by CDC 28 and state health departments, during January 1 -July 31, 2016 (Walker et al. 2016). In 29 July 2016 the first locally acquired cases of Zika virus (ZIKAV) from mosquitoes were 30 confirmed for the U.S. state of Florida (Likos, 2016). Aedes mosquitoes transmit ZIKAV, Americas (Guerbois et al. 2016, Ferreira-de-Brito et al. 2016. This is likely the result of 36 Ae. aegypti preferring to feed more frequently on humans (Scott et al. 1993(Scott et al. , 2000, and 37 being highly peridomestic compared to Ae. albopictus, which can inhabit more rural 38 environments (Braks et al. 2003;Tsuda et al. 2006). 39 In this study, we concentrated on U.S. Department of Defense (DoD) facilities but 40 the approach could be used for any area of interest. Some military facilities have long 41 standing mosquito surveillance programs (Foley et al. 2011a), and Zika virus capable of carrying ZIKAV occur, and increased vector monitoring will be conducted in 48 installations in 27 states, the District of Columbia, Guam and Puerto Rico (Kime, 2016). Public Health Center and regional Navy Environmental and Preventative Medicine 53 3 Units, and the Navy Entomology Center of Excellence, assist those undertaking vector 54 surveillance or arbovirus testing. 55 For a military entomologist tasked with establishing and maintaining an Aedes 56 spp. / ZIKAV surveillance program in temperate areas that experience high mosquito 57 seasonality, two important questions arise: 1) is ZIKAV transmission possible here?; 58 and 2) when should we conduct vector surveillance?. In the following we describe an 59 Excel-based tool that is designed to assist entomologists and other health personnel 60 address these two questions.

61
Habitat suitability models displaying potential distribution have been published for 62 both Ae. aegypti and Ae. albopictus (Attaway et al. 2016, Brady et al. 2014, Campbell et 63 al. 2015, Khormi & Kumar 2014, Medley 2010, as well as for ZIKAV (Carlson et al. 64 2016, Messina et al. 2016, Perkins et al. 2016, Samy et al. 2016. While these models 65 often display average yearly suitability they do not necessarily provide information that  Relative humidity, rainfall, drought, and wind velocity affect survival and behavior 72 of mosquitoes, and therefore transmission (Kramer & Ebel, 2003). However, 73 temperature is the most important ecological determinant of development rate in Ae. 74 aegypti (Couret & Benedict 2014), and one of the principal determinants of Aedes 75 survival (Brady et al. 2013). Temperature also directly affects the replication rate of 76 arboviruses, thus affecting the extrinsic incubation period (Gubler et al. 2007 (Macfie, 1920;Bliss & Gill, 1933;Christopher, 1960). In a 84 4 study of Ae. aegypti distribution using the program CLIMEX, Khormi & Kumar (2014) set 85 the limiting low temperature at 18 °C, the lower optimal temperature at 25 °C, the upper 86 optimal temperature at 32 °C and the limiting high temperature at 38 °C. Brady et al. 87 (2014) limited their predictions of temperature suitability to areas with a maximum 88 monthly temperature exceeding 13°C for Ae. albopictus and 14°C for Ae. aegypti. 89 These threshold temperatures were based on previous studies of the observed 90 temperatures below which biting and movement behaviors are impaired [Christophers, 91 1960;Estrada-Franco & Craig, 1995;Carrington et al. 2013a,b).

92
Studies suggest that an increase between 14-18 °C and 35-40 °C can lead to 93 higher transmission of dengue (Wallis, 2005). Xiao et al. (2014) found that oral 94 infections of DENV2 did not produce antigens in the salivary glands of Ae. albopictus 95 kept at 18°C for up to 25 days but did produce antigens at 21°C during this period. It is 96 not known if Ae. albopictus held longer at the lower temperature would have 97 disseminated infections, but Dohm et al. (2002) found that Culex pipiens required 25 98 days at 18°C to disseminate infections of West Nile Virus. For comparison, WNV is 99 capable of replication from 14-45°C (Cornel et al. 1993, Kinney et al. 2006). Tilston et that ZIKAV is more thermally stable than DENV, and is also structurally stable even 104 when incubated at 40°C, mimicking the body temperature of extremely feverish patients 105 after virus infection (but see Goo et al. 2016).

106
Remotely sensed temperature data is freely available from multiple sources as   We chose the models of Ae. aegypti and Ae. albopictus by Kraemer et al. (2015), 144 as these are recent and are based on an extensively documented set of presence 145 observations for each vector. For this study, we used the habitat suitability model for 146 ZIKAV transmission by Messina et al. (2016). The 0.5 model suitability score was 147 arbitrarily used as the cut-off for presence/absence.

347
It is important to note that each data source used in this analysis has the 348 potential for errors which should be considered when determining risk. For example, 349 habitat suitability models for each vector may not be accurate for all areas, and only 350 predict average yearly suitability. Temperature data refers to the maximum day-time air 351 temperature near the surface (averaged over various spatial resolutions) from daily data 352 for a recent date range, which NASA acknowledges has limitations. Vectors can also 353 seek microclimates (e.g. indoors, subterranean habitats) that may be warmer or cooler 354 than the outside temperature that is estimated by remote sensing data. Temperatures