Detection of bluetongue virus RNA in field-collected Culicoides spp. (Diptera: Ceratopogonidae) following the discovery of bluetongue virus serotype 1 in white-tailed deer and cattle in Louisiana

From forests to fields: investigating Culicoides (Diptera: Ceratopogonidae) abundance and diversity in cattle pastures and adjacent woodlands

Culicoides Latreille (Diptera: Ceratopogonidae) biting midges are hematophagous flies that feed on wild and domestic ruminants. They can transmit arboviruses, such as bluetongue virus (BTV) and epizootic hemorrhagic disease virus (EHDV), which circulate in the United States. Larvae occupy a range of aquatic and semiaquatic habitats, and disperse short distances from their development sites. In the southeastern United States, there are limited studies on the abundance and diversity of Culicoides in wooded and adjacent livestock pasture habitats. In this study, we characterized Culicoides diversity and abundance within these distinct habitat types. BG-Sentinel and CDC miniature suction traps baited with CO2 or UV-light were placed in wooded and pasture habitats at 2 locations on a university beef farm in Savoy, Arkansas. Traps were set once per week for 9 wk during August-October of 2021 and 2022. Fifteen species were collected during this study, and the 2 most abundant species were Culicoides haematopotus Malloch and Culicoides stellifer Coquillett. There was a significant effect of site and location on C. haematopotus collections, and a significant effect and interaction of site and trap on C. stellifer collections. In the woods, significantly more C. stellifer were collected from CDC-UV traps, while in the pasture significantly more were collected in CDC-CO2 traps. These data suggest that C. stellifer, a putative vector of BTV and EHDV in the southeast, may be traveling into the pasture to host-seek, while C. haematopotus remains primarily in wooded areas. This study reveals community differences between these habitat types and implications for Culicoides control.

Assessment of light-emitting diodes for sampling phlebotomines (Diptera: Psychodidae) from an urban park of the Brazilian Amazon

The present study aimed to assess different light sources for sampling phlebotomines (Diptera: Psychodidae) from Bosque Rodrigues Alves, a forested park surrounded by the urban area of Belém in the Brazilian Amazon. Centers for Disease Control traps, baited with blue, green, and warm white light-emitting diodes (LEDs), as test group, and incandescent light, as control group, were used. The electromagnetic spectra and luminous intensities of the light sources were characterized. Fractional vegetation cover at each sampling site was also estimated. Abundance, richness, rarefaction curves, Shannon and Simpson diversity indices, phlebotomines/trap/hour, and phlebotomines/trap/night were estimated and compared. The light sources of the test group presented greater luminous intensity than the control, but were similar to each other. There were no differences in vegetation cover at each site. A total of 1,346 phlebotomines comprising 11 species were sampled. The most abundant species were as follows: Nyssomyia antunesi (Coutinho, 1939), Trichophoromyia ubiquitalis (Mangabeira, 1942), Bichromomyia flaviscutellata (Mangabeira, 1942), and Th. brachipyga (Mangabeira, 1942). Light traps with LEDs had richness, abundance, and Shannon diversity indices similar to those obtained with incandescent light. The warm white LED had a higher Simpson's index than the other light sources. Phlebotomine responses to incandescent light were similar to those to LEDs in most analyses, confirming the applicability of these light sources as alternative devices for entomological surveillance. Low consumption ensures greater autonomy of the traps, providing better operability during fieldwork.

The Study of Bluetongue Virus (BTV) and Epizootic Hemorrhagic Disease Virus (EHDV) Circulation and Vectors at the Municipal Parks and Zoobotanical Foundation of Belo Horizonte, Minas Gerais, Brazil (FPMZB-BH)

Bluetongue Virus (BTV) and Epizootic Hemorrhagic Disease Virus (EHDV) are Orbiviruses primarily transmitted by their biological vector, Culicoides spp. Latreille, 1809 (Diptera: Ceratopogonidae). These viruses can infect a diverse range of vertebrate hosts, leading to disease outbreaks in domestic and wild ruminants worldwide. This study, conducted at the Belo Horizonte Municipal Parks and Zoobotany Foundation (FPMZB-BH), Minas Gerais, Brazil, focused on Orbivirus and its vectors. Collections of Culicoides spp. were carried out at the FPMZB-BH from 9 December 2021 to 18 November 2022. A higher prevalence of these insects was observed during the summer months, especially in February. Factors such as elevated temperatures, high humidity, fecal accumulation, and proximity to large animals, like camels and elephants, were associated with increased Culicoides capture. Among the identified Culicoides spp. species, Culicoides insignis Lutz, 1913, constituted 75%, and Culicoides pusillus Lutz, 1913, 6% of the collected midges, both described as competent vectors for Orbivirus transmission. Additionally, a previously unreported species in Minas Gerais, Culicoides debilipalpis Lutz, 1913, was identified, also suspected of being a transmitter of these Orbiviruses. The feeding preferences of some Culicoides species were analyzed, revealing that C. insignis feeds on deer, Red deer (Cervus elaphus) and European fallow deer (Dama dama). Different Culicoides spp. were also identified feeding on humans, raising concerns about the potential transmission of arboviruses at the site. In parallel, 72 serum samples from 14 susceptible species, including various Cervids, collected between 2012 and 2022 from the FPMZB-BH serum bank, underwent Agar Gel Immunodiffusion (AGID) testing for BTV and EHDV. The results showed 75% seropositivity for BTV and 19% for EHDV. Post-testing analysis revealed variations in antibody presence against BTV in a tapir and a fallow deer and against EHDV in a gemsbok across different years. These studies confirm the presence of BTV and EHDV vectors, along with potential virus circulation in the zoo. Consequently, implementing control measures is essential to prevent susceptible species from becoming infected and developing clinical diseases.

Host Associations of Culicoides Biting Midges in Northeastern Kansas, USA

Culicoides biting midges (Diptera: Ceratopogonidae) are hematophagous flies that transmit several viruses of veterinary concern to livestock. Understanding blood feeding behaviors is integral towards identification of putative vector species and preventing the transmission of these pathogens. PCR-based blood meal analysis was conducted on 440 blood-engorged Culicoides midges collected in northeastern Kansas, with 316 (71.8%) returning non-human vertebrate identifications at the ≥95% identity match level. Broadly, Culicoides sonorensis, Culicoides stellifer, and Culicoides variipennis were found to feed heavily on mammalian hosts, while Culicoides crepuscularis and Culicoides haematopotus fed on avian hosts. The blood meals in all specimens were graded prior to DNA extraction to determine whether blood meal size or digestion status significantly impacted the likelihood of a quality host match. Size had a significant impact on the likelihood of a quality match at grades 3-5, whereas digestion only significantly impacted outcomes at the most extreme grade. These vector-host dynamics have not previously been studied in Culicoides collected in Kansas, which represents a unique tallgrass prairie biome within the United States that is heavily interspersed with livestock operations. Based on these data, the highly abundant species C. crepuscularis and C. haematopotus are unlikely to be major vectors of mammalian viruses.

An updated catalogue of biting midges of the genus Culicoides Latreille, 1809 (Diptera, Ceratopogonidae) of Mexico and their known distribution by state

An updated catalogue of Culicoides of Mexico is presented. It includes 86 species with their regional distribution and corresponding record references, known immature stages and associated pathogens. In addition, a taxonomic key for subgenera and species groups for Mexico is presented and an index of species by state is included.

Assigning Culicoides larvae to species using DNA barcoding of adult females and phylogenetic associations

Background

Orbivirus-induced hemorrhagic diseases cause high mortality in wild and captive white-tailed deer in North America. The role of different Culicoides species in Orbivirus transmission outside of areas of intensive animal production has not been established. At our study location, bluetongue virus (BTV) RNA-positive female Culicoides debilipalpis pools have been detected annually since 2012 when BTV transmission was noted in a captive deer herd. Identifying specific larval habitats of suspected vectors at active transmission sites is crucial both for identifying the source of the vectors and for subsequently planning intervention actions. Since C. debilipalpis larvae are known to develop in tree holes, this study was designed to use DNA barcoding to identify larvae collected from tree holes.

Methods

Adult female Culicoides were collected using light or emergence traps and morphologically identified to 11 species. Culicoides sonorensis were also obtained from a laboratory colony. Substrate was collected from tree holes and flooded with water to harvest floating larvae. Total DNA from three to seven adult females per species and 19 larvae was extracted. Two loci of the nuclear 18S ribosomal RNA (rRNA) gene, one locus each of the mitochondrial cytochrome oxidase subunit I (COI) gene and the nuclear 28S rRNA gene were amplified using loci-specific primers.

Results

All 61 adults were sequenced at each of the four loci under study. Since no single locus delineated all putative species and the COI locus yielded unreliable pseudogenes for two individuals of C. arboricola, sequences of all four loci were concatenated to maximize species separation and allow for larval association with identified adults. Sixteen larvae were clearly assigned to species based on DNA barcoding and phylogenetic results. Multiple larvae were assigned to each of the C. debilipalpis clade, the C. villosipennis clade, the C. arboricola clade and the C. nanus clade.

Conclusions

Of the approximately 62 species described in the southeast USA, 21 have now been barcoded and sequences are publicly available. In this study, we constructed a database composed of species-specific sequences of adult Culicoides and then identified larvae to species by matching their corresponding sequences with adults. Since Culicoides larvae are difficult to identify, using DNA barcoding to facilitate larval habitat surveys can be a valuable tool.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s13071-022-05479-1.

Comparison of trapping methods for use in surveys for potential Culicoides vectors of orbiviruses

BACKGROUND

Bluetongue virus (BTV) and epizootic hemorrhagic disease virus (EHDV) are orbiviruses that can cause fatal vector-borne diseases in white-tailed deer (Odocoileus virginianus). Trapping methods for collecting potential Culicoides vectors of orbiviruses were compared to optimize surveillance studies.

METHODS

The number of captured midges and the virus infection rates of midge pools were compared for dry ice-baited Centers for Disease Control and Prevention (CDC) traps with or without black light. The number of individual midges of different Culicoides species captured at different crepuscular and nocturnal periods using rotator traps also was determined. The number of species/specimens of Culicoides was measured using five different trap methods including three animal-baited methods, a CDC trap with black light, and a CDC trap with no light.

RESULTS

In trial one, there was no significant difference (P = 0.37) in the proportion of BTV-infected flies caught in traps with light compared to traps without light. However, there was a significant difference (P = 0.026) for EHDV-infected flies, and 89% were captured in traps with light. In trial two, more specimens of C. debilipalpis were captured in the morning hours (06:00-08:00) than in the evening hours (18:00-20:00). For trial three, the animal-baited traps did not capture any species of Culicoides that were not captured in the CDC light traps. There was no significant difference (P = 0.22) in total specimens captured among all five trap types.

CONCLUSIONS

Specimens of Culicoides infected with BTV were not repelled by light traps in the first trial, while the majority of the specimens positive for EHDV were caught in traps with light. For the second trial, specimens of C. debilipalpis were most abundant during early morning hours, and thus spray applications of insecticides for control of that species may be more effective at sunrise rather than sunset. For objective three, no animal-baited trapping method collected different species of midges when compared to the CDC traps with light, which is unlike certain studies conducted in other geographical regions.

Detection of Vesicular Stomatitis Virus Indiana from Insects Collected during the 2020 Outbreak in Kansas, USA

Vesicular stomatitis (VS) is a reportable viral disease which affects horses, cattle, and pigs in the Americas. Outbreaks of vesicular stomatitis virus New Jersey serotype (VSV-NJ) in the United States typically occur on a 5-10-year cycle, usually affecting western and southwestern states. In 2019-2020, an outbreak of VSV Indiana serotype (VSV-IN) extended eastward into the states of Kansas and Missouri for the first time in several decades, leading to 101 confirmed premises in Kansas and 37 confirmed premises in Missouri. In order to investigate which vector species contributed to the outbreak in Kansas, we conducted insect surveillance at two farms that experienced confirmed VSV-positive cases, one each in Riley County and Franklin County. Centers for Disease Control and Prevention miniature light traps were used to collect biting flies on the premises. Two genera of known VSV vectors, Culicoides biting midges and Simulium black flies, were identified to species, pooled by species, sex, reproductive status, and collection site, and tested for the presence of VSV-IN RNA by RT-qPCR. In total, eight positive pools were detected from Culicoides sonorensis (1), Culicoides stellifer (3), Culicoides variipennis (1), and Simulium meridionale (3). The C. sonorensis- and C. variipennis-positive pools were from nulliparous individuals, possibly indicating transovarial or venereal transmission as the source of virus. This is the first report of VSV-IN in field caught C. stellifer and the first report of either serotype in S. meridionale near outbreak premises. These results improve our understanding of the role midges and black flies play in VSV epidemiology in the United States and broadens the scope of vector species for targeted surveillance and control.

Tracking Community Timing: Pattern and Determinants of Seasonality in Culicoides (Diptera: Ceratopogonidae) in Northern Florida

Community dynamics are embedded in hierarchical spatial–temporal scales that connect environmental drivers with species assembly processes. Culicoides species are hematophagous arthropod vectors of orbiviruses that impact wild and domestic ruminants. A better sense of Culicoides dynamics over time is important because sympatric species can lengthen the seasonality of virus transmission. We tested a putative departure from the four seasons calendar in the phenology of Culicoides and the vector subassemblage in the Florida panhandle. Two years of weekly abundance data, temporal scales, persistence and environmental thresholds were analyzed using a tripartite Culicoides β-diversity based modeling approach. Culicoides phenology followed a two-season regime and was explained by stream flow and temperature, but not rainfall. Species richness fit a nested pattern where the species recruitment was maximized during spring months. Midges were active year-round, and two suspected vectors species, Culicoides venustus and Culicoides stellifer, were able to sustain and connect the seasonal modules. Persistence suggests that Orbivirus maintenance does not rely on overwintering and that viruses are maintained year-round, with the seasonal dynamics resembling subtropical Culicoides communities with temporal-overlapping between multivoltine species. Viewing Culicoides-borne orbiviruses as a time-sensitive community-based issue, our results help to recommend when management operations should be delivered.

Prospective Study of Epizootic Hemorrhagic Disease Virus and Bluetongue Virus Transmission in Captive Ruminants

Bluetongue virus (BTV) and epizootic hemorrhagic disease virus (EHDV) cause hemorrhagic disease (HD) in wild ruminants and bluetongue disease (BT) and epizootic hemorrhagic disease (EHD) in livestock. These viruses are transmitted by biting midges in the genus Culicoides (family Ceratopogonidae). Mortality from this disease can reach 90% in certain breeds of sheep and in white-tailed deer (Odocoileus virginianus). From January until December of 2012, we conducted a prospective study to determine the origin and routes of transmission of BTV and EHDV in captive deer and cattle. The objective was to determine the abundance of Culicoides spp. and BTV/EHDV infection prevalence in midges, cattle, and deer in an area experiencing an outbreak of BT and EHD. Agar gel immunodiffusion (AGID) tests to detect for EHDV and BTV antibodies were conducted on serum collected from cattle and deer, quantitative reverse transcriptase polymerase chain reaction (RT-qPCR) was utilized for BTV/EHDV RNA detection in tissues from dead deer, and CDC miniature black light traps baited with dry ice were deployed to capture insects. The AGID results showed 19 out of 29 cattle and 18 out of 58 white-tailed deer seroconverted for these viruses during the vector season. Tradition gel-based reverse transcriptase polymerase chain reaction was utilized to determine serotype. Sixteen cows were positive for EHDV-2, EHDV-6, or BTV-12 and 15 deer positive for EHDV-1, EHDV-6, or BTV-12. Specimens from 14 species of Culicoides (Dptera: Ceratopogonidae) (Culicoides arboricola Root and Hoffman, Culicoides biguttatus Coquillett, Culicoides crepuscularis Malloch, Culicoides debilipalpis Lutz, Culicoides furens Poey, Culicoides haematopotus Malloch, Culicoides hinmani Khalaf, Culicoides nanus Root and Hoffman, Culicoides neopulicaris Wirth, Culicoides paraensis Goeldi, Culicoides stellifer Coquillet, Culicoides variipennis Coquillet, Culicoides villosipennis Root and Hoffman, and Culicoides venustus Hoffman) were captured and tested for BTV and EHDV using RT-qPCR assays. BTV viral nucleic acid was detected in three pools from three different species of midges: C. crepuscularis, C. debilipalpis, and C. stellifer.

Habitat associations of Culicoides species (Diptera: Ceratopogonidae) abundant on a commercial cervid farm in Florida, USA

BACKGROUND

Biting midges in the genus Culicoides (Diptera: Ceratopogonidae) transmit bluetongue virus (BTV) and epizootic hemorrhagic disease virus (EHDV) to ruminants, thus exerting a significant economic impact on animal agriculture worldwide. However, very little is known about the larval habitat characteristics of Culicoides species associated with BTV/EHDV transmission, particularly in southeastern USA, limiting the establishment of effective midge control strategies. In this study, we examined the habitat associations of Culicoides species abundant on a commercial cervid farm in Florida, USA and quantified several environmental variables of their habitat to identify the key variables associated with midge abundance.

METHODS

Mud/substrate samples from three potential larval habitats on the farm (edges of streams, puddles and seepages) were brought to the laboratory and incubated for adult emergence, and the percentage organic matter, macronutrients, micronutrients, pH, electrical conductivity, moisture and microbial concentrations of the substrate were quantified.

RESULTS

Strong habitat associations were observed for Culicoides haematopotus (Malloch) (stream edge), Culicoides stellifer (Coquillett) (puddles) and Culicoides loisae (Jamnback) (stream edge), the most commonly emerging midge species from the samples. Suspected vector species of BTV/EHDV on the property, C. stellifer and Culicoides venustus (Hoffman), emerged mainly from habitats with moderate-high levels of pollution (edges of puddles and seepages) as indicated by the relatively higher concentrations/levels of organic matter, nutrients and other environmental variables in these samples. The emergence of C. insignis was too low to form any meaningful conclusions. For each Culicoides species, only weak positive or negative associations were detected between midge abundance and the various environmental variables quantified.

CONCLUSIONS

Habitat associations of Culicoides species abundant on a local cervid/animal farm vary, most likely as a function of certain biotic/abiotic characteristics of the habitat. Further studies across a larger spatial and temporal scale will be needed to experimentally evaluate/identify the key factors more strongly associated with the abundance of target Culicoides species. This information, in the long term, can be potentially exploited to render local habitats unsuitable for midge oviposition/larval development.

Bluetongue virus infection creates light averse Culicoides vectors and serious errors in transmission risk estimates

BACKGROUND

Pathogen manipulation of host behavior can greatly impact vector-borne disease transmission, but almost no attention has been paid to how it affects disease surveillance. Bluetongue virus (BTV), transmitted by Culicoides biting midges, is a serious disease of ruminant livestock that can cause high morbidity and mortality and significant economic losses. Worldwide, the majority of surveillance for Culicoides to assess BTV transmission risk is done using UV-light traps. Here we show that field infection rates of BTV are significantly lower in midge vectors collected using traps baited with UV light versus a host cue (CO2).

METHODS

We collected Culicoides sonorensis midges in suction traps baited with CO2, UV-light, or CO2 + UV on three dairies in southern California to assess differences in the resulting estimated infection rates from these collections. Pools of midges were tested for BTV by qRT-PCR, and maximum likelihood estimates of infection rate were calculated by trap. Infection rate estimates were also calculated by trapping site within a dairy. Colonized C. sonorensis were orally infected with BTV, and infection of the structures of the compound eye was examined using structured illumination microscopy.

RESULTS

UV traps failed entirely to detect virus both early and late in the transmission season, and underestimated virus prevalence by as much as 8.5-fold. CO2 + UV traps also had significantly lower infection rates than CO2-only traps, suggesting that light may repel infected vectors. We found very high virus levels in the eyes of infected midges, possibly causing altered vision or light perception. Collecting location also greatly impacts our perception of virus activity.

CONCLUSIONS

Because the majority of global vector surveillance for bluetongue uses only light-trapping, transmission risk estimates based on these collections are likely severely understated. Where national surveillance programs exist, alternatives to light-trapping should be considered. More broadly, disseminated infections of many arboviruses include infections in vectors' eyes and nervous tissues, and this may be causing unanticipated behavioral effects. Field demonstrations of pathogen-induced changes in vector behavior are quite rare, but should be studied in more systems to accurately predict vector-borne disease transmission.

Measurement of the infection and dissemination of bluetongue virus in culicoides biting midges using a semi-quantitative rt-PCR assay and isolation of infectious virus

Background

Culicoides biting midges (Diptera: Ceratopogonidae) are the biological vectors of globally significant arboviruses of livestock including bluetongue virus (BTV), African horse sickness virus (AHSV) and the recently emerging Schmallenberg virus (SBV). From 2006–2009 outbreaks of BTV in northern Europe inflicted major disruption and economic losses to farmers and several attempts were made to implicate Palaearctic Culicoides species as vectors. Results from these studies were difficult to interpret as they used semi-quantitative RT-PCR (sqPCR) assays as the major diagnostic tool, a technique that had not been validated for use in this role. In this study we validate the use of these assays by carrying out time-series detection of BTV RNA in two colony species of Culicoides and compare the results with the more traditional isolation of infectious BTV on cell culture.

Methodology/Principal Findings

A BTV serotype 1 strain mixed with horse blood was fed to several hundred individuals of Culicoides sonorensis (Wirth & Jones) and C. nubeculosus (Mg.) using a membrane-based assay and replete individuals were then incubated at 25°C. At daily intervals 25 Culicoides of each species were removed from incubation, homogenised and BTV quantified in each individual using sqPCR (C_q values) and virus isolation on a KC-C. sonorensis embryonic cell line, followed by antigen enzyme-linked immunosorbent assay (ELISA). In addition, comparisons were also drawn between the results obtained with whole C. sonorensis and with individually dissected individuals to determine the level of BTV dissemination.

Conclusions/Significance

C_q values generated from time-series infection experiments in both C. sonorensis and C. nubeculosus confirmed previous studies that relied upon the isolation and detection of infectious BTV. Implications on the testing of field-collected Culicoides as potential virus vectors by PCR assays and the use of such assays as front-line tools for use in diagnostic laboratories in this role are discussed.

Record of two species of Culicoides (Diptera, Ceratopogonidae) new for Madagascar and molecular study showing the paraphylies of the subgenus Oecacta and the Schultzei group

Culicoides are vectors of diseases of Veterinary Medicine importance (bluetongue, African horse sickness, Schmallenberg virus) all over the world. In the present study, we report two species new for Madagascar: C. nevilli and C. enderleini. They belong to the Schultzei group which is sometimes classified in the subgenus Oecacta and sometimes in the subgenus Remmia, depending on authors. Consequently, we carried out a molecular cladistics of these groups based on cytochrome C oxidase subunit I mtDNA sequences. We processed the Malagasy specimens and some C. furens (the Oecacta type-species) caught in Florida and we analyzed their sequences and those available in Genbank: C. schultzei, C. oxystoma, C. festivipennis, C. brunnicans, C. kibunensis, C. truncorum and C. vexans. C. (Avaritia) imicola have been selected as an outgroup. The maximum parsimony analysis showed the paraphylies of the Schultzei group (=Remmia) and of the subgenus Oecacta if the first group is excluded from the latter. Our results underline the doubtful current classification and need to be validated by other molecular markers in the future.

Implicating Culicoides biting midges as vectors of Schmallenberg virus using semi-quantitative RT-PCR

BACKGROUND

The recent unprecedented emergence of arboviruses transmitted by Culicoides biting midges in northern Europe has necessitated the development of techniques to differentiate competent vector species. At present these techniques are entirely reliant upon interpretation of semi-quantitative RT-PCR (sqPCR) data in the form of Cq values used to infer the presence of viral RNA in samples.

METHODOLOGY/PRINCIPAL FINDINGS

This study investigates the advantages and limitations of sqPCR in this role by comparing infection and dissemination rates of Schmallenberg virus (SBV) in two colony lines of Culicoides. Through the use of these behaviorally malleable lines we provide tools for demarcating arbovirus infection and dissemination rates in Culicoides which to date have prevented clear implication of primary vector species in northern Europe. The study demonstrates biological transmission of SBV in an arthropod vector, supporting the conclusions from field-caught Culicoides and provides a general framework for future assessment of vector competence of Culicoides for arboviruses using sqPCR.

CONCLUSIONS/SIGNIFICANCE

When adopting novel diagnostic technologies, correctly implicating vectors of arboviral pathogens requires a coherent laboratory framework to fully understand the implications of results produced in the field. This study illustrates these difficulties and provides a full examination of sqPCR in this role for the Culicoides-arbovirus system.

Spatial-temporal model of haemorrhagic disease in white-tailed deer in south-east USA, 1983 to 2000

The present study constructed a spatial-temporal statistical model to identify the risk and protective factors for haemorrhagic disease (HD) in white-tailed deer in the five states of Alabama, Georgia, South Carolina, North Carolina and Tennessee. The response variable was binary, indicating the presence or absence of HD in an individual county, measured annually from 1983 to 2000. Predictor variables included climatic factors of temperature, rainfall, wind speed and dew point, remotely sensed data of normalised difference vegetation index (NDVI) and land surface temperature derived from archived remotely sensed advanced very-high-resolution radiometer (AVHRR) satellite data, elevation, a spatial autocorrelation (SA) term and a temporal autocorrelation term. This study first applied principal component factor analysis to reduce the volume of climatic data and remotely sensed data. Then, a generalised linear mixed model framework (GLMM) was used to develop a spatial-temporal statistical model. The results showed that the area under receiver operating characteristic curve (ROC) was 0.728, indicating a good overall fit of the model. The total prediction accuracy over the 18 year period with optimal cut-off probability was 67 per cent. The prediction accuracy for individual years ranged from 48 to 75 per cent.