ECO‐EPIDEMIOLOGY OF DENGUE AND ROSS
RIVER VIRUSES ACROSS RURAL AND URBAN
ENVIRONMENTS
Amanda Murphy BSc (Hons), MIPH
Submitted in fulfilment of the requirements for the degree of Doctor of Philosophy
School of Biomedical Sciences Faculty of Health Queensland University of Technology 2020
Statement of Original Authorship
The work contained in this thesis was undertaken between QUT and QIMR Berghofer Medical Research Institute, and has not been previously submitted to meet requirements for an award at any other higher education institution. To the best of my knowledge and belief, the thesis contains no material previously published or written by another person except where due reference is made.
Signature: QUT Verified Signature
Date: ____12 September, 2020____
Chapter 1: Introduction 2
Table of Contents
Statement of Original Authorship ...... 2 Table of Contents ...... 3 Supervisory Team ...... 5 List of publications arising from this research ...... 6 List of Tables ...... 7 List of Figures ...... 8 List of Supplementary Tables and Figures ...... 10 Keywords ...... 11 List of Abbreviations ...... 12 Acknowledgements ...... 13 Summary ...... 16
CHAPTER 1: INTRODUCTION ...... 19 Background ...... 19 Context ...... 20 Thesis Outline ...... 22
CHAPTER 2: LITERATURE REVIEW & RESEARCH OBJECTIVES ...... 24 Global burden of arboviral disease ...... 24 Dengue epidemiology ...... 27 Ross River virus epidemiology ...... 30 Prevention and control of arboviral disease ...... 37 Summary and implications ...... 40 Research objectives ...... 43 Significance and Scope ...... 44
CHAPTER 3: RESULTS FOR OBJECTIVE 1 ...... 46 Statement of Contribution of Co‐Authors for Thesis by Published Paper ...... 47 3.1. Incidence and epidemiologic features of dengue in Sabah, Malaysia ... 48 3.2. Additional analyses of dengue data ...... 87
Chapter 1: Introduction 3 CHAPTER 4: RESULTS FOR OBJECTIVE 2 ...... 90 Statement of Contribution of Co‐Authors for Thesis by Published Paper ...... 92 4.1. Spatial and temporal patterns of Ross River virus in South East Queensland, Australia: identification of hot spots at the rural‐urban interface ...... 93
CHAPTER 5: RESULTS FOR OBJECTIVE 3 ...... 137 Statement of Contribution of Co‐Authors for Thesis by Published Paper ..... 139 5.1. A micro‐PRNT for the detection of Ross River virus antibodies in mosquito blood meals: a useful tool for inferring transmission pathways ...... 140
CHAPTER 6: RESULTS FOR OBJECTIVE 4 ...... 163 Statement of Contribution of Co‐Authors for Thesis by Published Paper ..... 165 6.1. Mosquitoes as flying syringes: using mosquito blood meals to implicate vectors and hosts in Ross River virus transmission ...... 166
CHAPTER 7: GENERAL DISCUSSION ...... 202
CHAPTER 8: REFERENCES ...... 212
CHAPTER 9: APPENDICES ...... 227 Appendix A: Interpreting mosquito feeding patterns in Australia through an ecological lens: an analysis of blood meal studies ...... 228 Appendix B: RRV field survey protocol ...... 260 Appendix C: Associations between Ross River virus infection in humans and vector‐vertebrate community ecology in Brisbane, Australia ...... 289
Chapter 1: Introduction 4
Supervisory Team
Associate Professor Greg Devine, Mosquito Control Laboratory, QIMR Berghofer Medical Research Institute (QIMRB) – Principal QIMRB Supervisor
Dr Francesca Frentiu, School of Biomedical Sciences, and Institute for Health and Biomedical Innovation, Queensland University of Technology – Principal QUT Supervisor
Professor Wenbiao Hu, School of Public Health and Social Work, Queensland University of Technology – QUT Associate Supervisor
Professor Louise Hafner, School of Biomedical Sciences, and Institute for Health and Biomedical Innovation, Queensland University of Technology – QUT Associate Supervisor
Dr Cassie Jansen, Communicable Diseases Branch, Queensland Health – External Associate Supervisor
Chapter 1: Introduction 5 List of publications arising from this research
Chapter 3: Murphy AK, Rajahram GS, Jilip J, Maluda M, William T, Hu W, Reid S, Devine GJ, Frentiu FD. Incidence and epidemiologic features of dengue in Sabah, Malaysia. PLoS Neglected Tropical Diseases, 11 May, 2020.
Chapter 4: Murphy AK, Clennon JA, Vazquez‐Prokopec G, Jansen CC, Frentiu FD, Hafner LM, Hu W, Devine GJ. Spatial and temporal patterns of Ross River virus in South East Queensland, Australia: identification of hot spots at the rural‐urban interface. Under review with BMC Infectious Diseases, submitted 9 March, 2020.
Chapter 5: Gyawali N, Murphy AK, Hugo LE, Devine GJ. A micro‐PRNT for the detection of Ross River virus antibodies in mosquito blood meals: a useful tool for inferring transmission pathways. PLoS One, 24 July, 2020.
Chapter 6: Murphy AK, Graham M, Gyawali N, Skinner EB, Jansen CC, Shivas MA, Onn MB, Rabellino A, Hu W, Hafner LM, Frentiu FD, Devine GJ. Mosquitoes as flying syringes: investigation of Ross River virus epidemiology and host seroprevalence using mosquito blood meals. In preparation.
Appendix A: Stephenson EB, Murphy AK, Jansen CC, Peel AJ, McCallum, H. Interpreting mosquito feeding patterns in Australia through an ecological lens: an analysis of blood meal studies. Parasites and Vectors, 4 April, 2019.
Appendix C: Stephenson EB, Murphy AK, Jansen CC, Shivas M, McCallum, H, Peel AJ. Associations between Ross River virus disease and vector and vertebrate community ecology in Brisbane, Australia. Vector‐Borne and Zoonotic Diseases, 4 May, 2020.
Chapter 1: Introduction 6
List of Tables
Table 1.1. Outline of manuscripts forming this thesis ...... 23
Table 2.1. Comparison of epidemiological characteristics of dengue and RRV...... 41
Table 3.1. Summary of population and dengue burden across Sabah, 2010‐ 2016...... 80
Table 3.2. Mosquito larvae species collected from case residences in Sabah during 2015‐2016...... 82
Table 4.1. Summary characteristics of each Local Government Area (LGA) of South East Queensland during the study period, 2001‐2016...... 120
Table 6.1. Adult mosquitoes captured across Brisbane...... 189
Table 6.2. Host origins of mosquito blood meals...... 192
Table 6.3. Seropositive hosts identified from mosquito blood meals...... 195
Chapter 1: Introduction 7 List of Figures
Figure 2.1. Transmission cycle of dengue...... 30
Figure 2.2. Transmission cycle of RRV...... 34
Figure 3.1. Map of Malaysia and Sabah state...... 75
Figure 3.2. Temporal pattern of dengue in Sabah, 2010‐2016...... 76
Figure 3.3. State‐wide annual incidence of dengue in rural and urban localities, 2011‐2016...... 77
Figure 3.4. Incidence of dengue in Sabah by age group and gender, 2010‐2016. .... 78
Figure 3.5. Annual spatial incidence of dengue in Sabah, 2010‐2016...... 79
Figure 3.6. Total severe dengue notifications by district, 2010‐2016...... 81
Figure 4.1. Map of Australia and South East Queensland...... 117
Figure 4.2. Mean annual RRV incidence rates by age‐group in South East Queensland, 2001‐2016...... 118
Figure 4.3. Mean monthly trend of RRV case notifications in South East Queensland, 2001‐2016...... 119
Figure 4.4. Mean annual RRV incidence of State Suburb Codes (SSCs) in South East Queensland, 2001‐2016...... 121
Figure 4.5. Mean annual RRV incidence of rural and urban State Suburb Codes (SSCs), 2001‐2016...... 122
Figure 4.6. Persistent high incidence hot spots in South East Queensland, 2001‐ 2016...... 123
Figure 5.1. Aedes notoscriptus at different stages of blood meal digestion...... 158
Figure 5.2. Plaque neutralisation demonstrated by standard PRNT or micro‐ PRNT using koala serum samples...... 159
Figure 5.3. The effect of mosquito homogenates on plaque formation...... 160
Chapter 1: Introduction 8
Figure 5.4. Percent neutralisation of RRV by vertebrate antibodies in mosquito blood meals harvested at different time points post blood‐feeding. ... 161
Figure 5.5. Impact of post‐feeding times on neutralisation of RRV...... 162
Figure 6.1. Map of Australia and Brisbane...... 188
Figure 6.2. Adult mosquitoes captured in urban and suburban sites...... 190
Figure 6.3. Relative abundance of vertebrate species across Brisbane sites...... 191
Figure 6.4. Host origins of mosquito blood meals by urban and suburban collection sites...... 193
Figure 6.5. RRV seropositive hosts in urban and suburban sites...... 194
Chapter 1: Introduction 9 List of Supplementary Tables and Figures
S3.1 Figure. Seasonal decomposition of incidence rates in Sabah, 2010‐2016...... 84
S3.2 Figure. Variation in dengue incidence across Sabah districts, 2010‐2016...... 85
S3.3 Table. Entomological surveillance of case residences by district, 2015‐ 2016...... 86
S3.4 Figure. Annual space‐time clusters of dengue in Sabah...... 88
S3.5 Table. Details of dengue space‐time cluster events in Sabah...... 89
S4.1 Figure. Monthly trend of RRV notifications in South East Queensland, 2001‐ 2016...... 125
S4.2 Table. Summary of annual case counts for each Local Government Area (LGA)...... 126
S4.3 Figure. Annual RRV incidence in South East Queensland: 2013‐2016...... 127
S4.4 Table. Locations with the highest overall rates across all years, 2001‐2016...... 128
S4.5 Figure. Annual hot and cold spots for RRV incidence in South East Queensland: 2013‐2016...... 131
S4.6 Figure. Persistent and mean RRV hot spots in South East Queensland: 2001‐2016...... 132
S4.7 Table. Summary of 45 persistent hot spots identified in both raw and smoothed incidence analyses from 2001‐2016...... 133
S6.1 Figure. Mosquitoes captured across Brisbane by month...... 197
S6.2 Table. Mosquitoes captured across Brisbane by each trap method...... 198
S6.3 Table. Host species origins of blood fed mosquitoes...... 199
S6.4 Table. Bird species origins of blood fed mosquitoes...... 200
S6.5 Table. RRV seroprevalence of mosquito blood meal hosts across urban and suburban sites...... 201
Chapter 1: Introduction 10
Keywords
Aedes albopictus, arbovirus, Australia, blood feeding, Borneo, dengue, ecology, epidemic, epidemiology, hosts, hot spot, landscapes, land use, mosquito, peri‐urban, Queensland, Ross River virus, rural, Sabah, seroprevalence, spatial, South East Asia, temporal, urban, vector‐borne disease, vectors, virus, xenodiagnosis, zoonosis.
Chapter 1: Introduction 11 List of Abbreviations
Abbreviation Definition
ABS Australian Bureau of Statistics AGSC Australian Geographical Standard Classification ALUM Australian Land Use Management BFV Barmah Forest virus CHIKV Chikungunya virus DENV Dengue virus LGA Local Government Area MARC Mosquito and Arboviral Reference Committee MVEV Murray Valley Encephalitis virus NoCS Notifiable Conditions System PCR Polymerase Chain Reaction PNG Papua New Guinea QIMRB QIMR Berghofer Medical Research Institute QUT Queensland University of Technology RRV Ross River virus SEQ South East Queensland SSC State Suburb Code WNV West Nile virus YFV Yellow fever virus ZIKV Zika virus
Chapter 1: Introduction 12
Acknowledgements I would like to thank the each of my supervisors for their hard work and valuable advice throughout my PhD. Special thanks go to my two primary supervisors, Greg Devine and Francesca Frentiu. Greg, thank you for your unwavering support throughout all my trials and triumphs; and Francesca, thank you for your valuable and broad‐ranging input and guidance. You both provided me with very different, yet complementary, support and I was lucky to have you. To my co‐supervisory team: thank you to Wenbiao Hu for your kindness and patience in guiding me through my analyses, to Cassie Jansen for your thoughtful input and enthusiasm, and Louise Hafner for your ongoing encouragement.
I thank my colleagues from the Mosquito Control Lab at QIMRB, and the Environmental Epidemiology Group at QUT, who provided regular help and support; in particular: Narayan Gyawali, Melissa Graham and Oselyne Ong for helping me survive lab work. Gordana Rasic, Igor Filipovic, Brian Johnson, Jon Darbro, Leon Hugo and others also provided useful input, along with welcome distractions. Thanks also go to the Disease Ecology Group at Emory University in Atlanta, USA, and specifically Gonzalo Vazquez‐Prokopec and Julie Clennon, who provided training in essential skills required for my research.
I am especially grateful for the time and effort contributed by the Mosquito Management Teams at Brisbane City Council, and other local governments: their valuable expertise and willingness to help with my field work was much appreciated. In particular, thank you Martin Shivas, Michael Muller, and the wonderful Michael Onn for your input, and for being so lovely to work with. I would have had a much harder time completing my field work if not for your support. Thank you also Eloise Skinner, my field work and PhD buddy, for your intellectual and moral support, constant energy and for being a good sounding board when needed.
For providing incredible personal support and encouragement throughout the whole PhD process, I thank my amazing life partner, Andrea Rabellino, and the following friends and family who contributed greatly to my motivation and wellbeing: Jennifer Lee, Rebekah McBride, Keith Rickhart, Jo Durham, Lisa and Craig Kelly, and Sukina and Effina Murphy‐Rabellino.
Chapter 1: Introduction 13 For supporting my annual progress reviews, additional thanks go to Katya Fischer (QIMRB), Patricia Dale (Griffith University), Helen Faddy and Elvina Viennet for their time and effort. Finally, I am grateful for financial and administrative resources provided by QIMRB, QUT, the Mosquito and Arbovirus Reference Committee, and the Australian Government who provided me a Research Training Program Scholarship.
Chapter 1: Introduction 14
There’s no such word as can’t.
Chapter 1: Introduction 15 Summary Arthropod‐borne viruses (arboviruses) are transmitted by the bite of infected arthropods, mainly mosquitoes and ticks. Once minor contributors to the global human disease burden, the world is experiencing a dramatic increase in transmission of arboviruses due to urbanisation, globalisation and the attendant movement of humans, vectors and pathogens. This research aimed to investigate the epidemiology and ecology of two arboviruses of public health importance: dengue virus (DENV) and Ross River virus (RRV). Patterns of disease caused by these viruses, and their influencing factors, were explored across different transmission settings. This entailed investigating epidemiological data for DENV and RRV across rural, peri‐urban and urban areas over space and time (Chapters 3 and 4). For DENV, I examined human case and vector surveillance data volunteered by a state public health department in Malaysia. For RRV, I analysed human case notifications along with prospectively collected field data of potential vector mosquitoes and vertebrate hosts.
I chose DENV because it causes the most prevalent arboviral disease in the world, dengue fever, with about 100 million symptomatic cases reported every year. Dengue fever has an economic burden of US$10 billion annually, and varying rates of mortality depending on access to treatment. The severe (haemorrhagic) form of the disease is a particular concern, and can require intensive care. In Chapter 3, I examined trends in reported dengue cases in the predominantly rural state of Sabah, in Malaysian Borneo – an area where sylvatic and urban circulation of pathogens intersect but where few empirical demographic, spatial or temporal analyses have been performed. Using a public health data set of dengue cases in Sabah between 2010 and 2016, I assessed patterns of disease and associated demographic, spatial and entomological risk factors. I found that dengue incidence rates were highest overall in the western districts of the state, while severe dengue cases were more focused in the east. Incidence rates were equivalent between urban and rural areas, and entomological collections indicated that the sylvatic vector Aedes albopictus was the most common potential vector present in the vicinity of both urban and rural dengue cases. The demographic group most at risk was males aged between 10‐29 years. Overall, I observed the magnitude of dengue outbreaks in Sabah to be increasing in both urban and rural settings, and suggest that
Chapter 1: Introduction 16
additional investigations are needed to better understand the drivers of risk in this understudied area.
In Chapter 4, I turned my attention to Ross River virus (RRV), a complex zoonosis that is poorly understood compared with DENV. RRV is responsible for the most commonly reported mosquito‐borne disease in Australia, and is associated with substantial annual morbidity and economic cost. The virus transmission dynamics vary with geographical and ecological environment, with an increasing trend in cases reported from urban areas. Although endemic to a range of bioclimatic zones across Australia and the Pacific, the highest disease prevalence is in Queensland. I focused my research in the highly populated south east region of the state, where I examined some of the underlying determinants of RRV epidemiology. I investigated spatial and temporal patterns of human RRV cases in detail across the suburbs of nine Local Government Areas with high rates of disease and found that spatial patterns were highly heterogeneous between outbreak years. Conversely, temporal patterns were highly predictable, following a distinct seasonal pattern. RRV disease was highly endemic across both urban and rural suburbs of the region, but the highest incidence and the most persistent hot spots were found in comparatively rural suburbs north of Brisbane. There was an overall trend for hot spots to be located along the peri‐urban edges of major urban areas, where urban residential and conserved natural landscapes intersect.
In the final two chapters (5 and 6), I attempted to further investigate the reasons for these disease trends by conducting field studies of RRV ecology. This involved surveys of both mosquitoes and wildlife from four urban and suburban study sites in Brisbane (Chapter 6). The field data were used together with a new laboratory xenodiagnostic technique – developed as part of this thesis – to assess wildlife host serostatus from blood‐fed mosquitoes (Chapter 5). This multifaceted approach allowed me to combine information on vectors, hosts and exposure to RRV. I was able to examine vector and host species composition between sites, as well as the implications of mosquito‐host feeding patterns in those sites. From capturing and analysing the contents of mosquito blood meals, I found that the most common vertebrate host fed on by mosquitoes in both urban and suburban sites was humans (76% of all blood meals were of human origin), with birds being the second most common (18%). I found a relatively low
Chapter 1: Introduction 17 proportion feeding on other vertebrates, including marsupials (6%). Further, mosquito blood meals tested for antibodies to RRV demonstrated prior RRV exposure in 20% of human blood meal samples, and 16% in blood meals of bird origin. The unique approach taken in these field and lab studies can help to infer potential transmission pathways between vector and reservoir hosts in different habitat types by simultaneously providing information on vector‐host feeding patterns and an estimation of virus exposure levels of wildlife hosts.
The outcomes of this research, discussed in Chapter 7, demonstrate the complexity and dynamism of dengue and RRV over their broad endemic ranges, and offers a more definitive analysis of transmission in those habitats than has been conducted previously. I demonstrate that dengue is equally burdensome in rural and urban areas of Sabah and that Aedes albopictus may play a key role in both. For RRV, I demonstrate how spatial and temporal analysis techniques, combined with comprehensive field studies and new xenodiagnostic tools, can inform our understanding of transmission pathways, public health risks and strategies for their mitigation. Broader outcomes of this research include increased understanding of the complex influences on arbovirus transmission in different human habitats, and the development of unique field and laboratory‐based approaches for investigating how the interactions of mosquitoes, wildlife hosts, humans and their environments can cause disease.
Chapter 1: Introduction 18
Chapter 1: Introduction
Background Vector‐borne diseases are a group of parasitic, bacterial and viral infections spread mainly by blood‐sucking insects, and comprise more than 17% off all reported infectious diseases globally [1]. This equates to billions of people worldwide at risk, and more than 700,000 reported deaths each year [2]. Survivors of vector‐borne diseases often suffer considerable physical, social and economic disability [2, 3]. The incidence and geographic distribution of vector‐borne diseases have increased significantly during the past few decades, stimulated by global increases in urban development, travel and trade, societal and environmental change, and by shifts in public health policy [4‐6]. These events have created novel opportunities for interactions between humans, animals, vectors and the disease‐causing pathogens they carry – resulting in emergence, or re‐emergence, of vector‐borne diseases [7, 8].
Prime examples are the recent resurgence of epidemic viruses spread by arthropods (arboviruses), notably dengue (DENV), chikungunya (CHIKV), Zika virus (ZIKV) and Yellow fever virus (YFV). These viruses, transmitted by the bite of infected mosquitoes, are a cause of rising global public health concern [8‐10]. Their introduction to new regions with non‐immune populations has produced sudden, large outbreaks that have overwhelmed health systems and caused considerable morbidity [11, 12]. Many of these outbreaks have been facilitated by the global spread of the Aedes (Ae.) vector species, Ae. aegypti (from African forests) and Ae. albopictus (from Asia) [10, 13‐ 15]. These species, which can act as vectors for multiple arboviruses, have been transported to tropical and temperate zones of all continents via human movement and trade [13‐16]. This consequently increased the spatial limits of the pathogens these vectors transmit, and the vulnerability of human populations to future arboviral emergence [12, 17].
The most prevalent and widespread global arboviral threat is DENV, for which over half the world’s population is at risk [18, 19]. Dengue disease incidence has been
Chapter 1: Introduction 19 increasing globally since the 1950’s, and expanding to previously unaffected countries [18, 20]. Other medically important arboviruses include CHIKV, which emerged in Europe (Italy) for the first time in 2007 [21, 22], and in Latin America and the Caribbean in 2013 [23, 24], followed by ZIKV which caused large‐scale epidemics in the Pacific Islands and Latin America from 2013 [23, 25]. Yellow fever virus caused an outbreak in Angola in 2016 that spread to Kenya, the Democratic Republic of the Congo, China, Nigeria and Brazil by 2019, instigating large vaccination campaigns [11, 26, 27]. These epidemics have necessitated substantial investment of public health resources [6, 11]. With continuing proliferation of global human movement and activity, there will be further opportunities for arboviral spread. [28‐30]. In order to devise strategies to limit the future burden of arboviral emergence and spread, the links between human activity and arboviral disease risk must be better defined.
Context
In this thesis, I present an investigation of two different arboviruses of public health importance which are both associated with geographic and ecological expansion: DENV and Ross River virus (RRV). Both viruses are transmitted across diverse habitats, with limited understanding of how the drivers of disease change between habitats, and incidence rates for both diseases are rising. Here, I examine recent disease trends for both viruses in urban and rural environments, and explore potential risk factors that might explain these trends.
I investigate DENV in a predominantly rural state of Malaysia, where incidence rates have risen in recent years. Dengue is an arbovirus of global importance, although the highest burden is in South East Asia [18]. This region has experienced frequent and large epidemics over the last 50 years, and yet effective responses to reduce infection risk are still not well characterised [31, 32]. These epidemics have been stimulated by the presence of large human populations in rapidly expanding urban environments that provide ideal conditions for breeding of the urban‐adapted vector, Ae. aegypti [20, 33, 34]. High levels of human movement through endemic areas also contributes, providing constant circulation of multiple DENV serotypes [35‐37]. Although now considered primarily a disease of urban environments of South East Asia, DENV also impacts rural
Chapter 1: Introduction 20
communities where it can circulate in sylvatic cycles that include primate hosts [38]. In rural environments, the vector Ae. albopictus is thought to be the dominant vector [39]; however, Ae. aegypti is also present in many rural locations, and vice versa [40‐43]. Similarly, sylvatic dengue strains can also be detected in urban environments [44]. Hence, the distinction between urban and rural transmission cycles is not always clear.
Similarly, RRV’s transmission cycle also traverses both urban and rural habitat types, though the drivers of infection are more poorly understood than for dengue [45]. The virus is endemic to Australia and some Pacific Island countries and territories, and has the highest reported incidence in Australia [46, 47]. Although historically associated with rural environments, RRV incidence has gradually increased in urbanised areas of Australia since the 1990s [48‐52]. Ross River virus circulation is maintained by multiple mosquito and animal reservoir species, with regular spillover to the human population, and its transmission pathways are challenging to disentangle [45, 49]. RRV’s natural wildlife reservoir is thought to be marsupials [49]; however, it is likely that the virus is maintained by additional (as yet unknown) hosts, particularly in urban areas [45, 53, 54]. Recent studies revealed that RRV also circulates more regularly than previously thought in Australia’s neighbouring Pacific Island countries including Fiji, French Polynesia and American Samoa, where marsupials are absent [55‐58]. This challenged some long‐held assumptions about RRV, and raised the possibility that the virus may be capable of expanding well outside of its presumed geographic range [59, 60]. My investigations of RRV focused in the state of Queensland, Australia, where there are persistently high rates of disease, and where the determinants of infection can be explored in a range of environments.
For both DENV and RRV, there is a need to more comprehensively dissect the influences on transmission in relation to environmental change. For the disease caused by each virus, there are no specific treatments, and hence prevention is the main form of disease control. Addressing the current and future impacts of these diseases requires greater understanding, monitoring and management of the underlying drivers.
Chapter 1: Introduction 21 Thesis Outline This thesis demonstrates the work undertaken in fulfilment of the IF49 Doctor of Philosophy degree and is presented as a series of manuscripts for publication. These are prefaced by a review of relevant literature, and are linked, interpreted and discussed by the Introduction (Chapter 1), Discussion (Chapter 7) and the inclusion of a preamble to each chapter that states its significance and how it contributes to the thesis. This thesis comprises 4 manuscripts: 3 first‐author and 1 second‐author. Contributions were also made to an additional two publications (as second author), which are not specifically included in this thesis but are included as Appendices, as they form part of another student’s thesis. All publications are outlined in Table 1.1, and in the List of publications arising from this research.
Chapter 1: Introduction 22
Table 1.1. Outline of manuscripts forming this thesis
Study type Chapter Manuscript title Focus
Description of disease notification trends, Incidence and Descriptive identification of high‐ epidemiologic features epidemiological 3.1 risk locations, of dengue in Sabah, study exploration of potential Malaysia environmental risk factors Spatial and temporal Description of disease patterns of Ross River notification trends, Descriptive virus in South East identification of high‐ epidemiological 4.1 Queensland, Australia: risk locations, study identification of hot exploration of potential spots at the rural‐urban environmental risk interface factors Proof of A micro‐PRNT for the Laboratory validation of concept study detection of Ross River micro‐PRNT for testing for a newly‐ virus antibodies in anti‐RRV antibody 5.1 developed mosquito blood meals: a status in wild animal laboratory useful tool for inferring hosts, using mosquito assay transmission pathways blood meals Mosquitoes as flying Molecular analysis of Exploratory syringes: investigation of host origins and RRV‐ field and Ross River virus 6.1 antibody status of field‐ laboratory epidemiology and host collected mosquito study seroprevalence using blood meals mosquito blood meals Interpreting mosquito feeding patterns in Review of existing Systematic Australia through an evidence of vector‐host review and 10: Appendix A ecological lens: an interactions and human meta‐analysis analysis of blood meal health implications studies Associations between Field study and Ross River virus disease Analysis of relationship analysis of RRV and vector and between RRV incidence 10: Appendix C vector‐host vertebrate community and vector‐host ecology ecology in Brisbane, composition Australia
Chapter 1: Introduction 23 Chapter 2: Literature review & Research
objectives
Global burden of arboviral disease
Arthropod‐borne virus infections are a global health issue of increasing importance [11]. Of more than 500 identified arboviruses, about 100 cause human disease, with a few (such as dengue, chikungunya, West Nile and Zika) presenting significant public health burden on a global scale [25, 61, 62]. Although arboviral pathogens have been causing human disease for centuries, emergence and spread of these viruses to new populations has intensified during the past 3 decades, causing substantial human morbidity and mortality on an international scale [5, 23]. There are substantial economic costs for countries with endemic disease in terms of vector control, case management and surveillance and prevention activities, as well as income and productivity losses associated with illness, treatment‐seeking and caregiving [2, 63].
Historically, arboviruses existed primarily in sylvatic transmission cycles between wildlife vertebrate and insect species, and were more localised to particular countries or regions [4]. Humans were an incidental host when they, or infected vectors, moved outside of their usual ecological range [64, 65]. However, the processes of globalisation, including expansion of travel, trade, urbanisation and human movement have provided opportunities for arboviruses and their vectors to expand their geographical and ecological range [6, 12, 64]. The ability of arboviruses and their vectors to opportunistically exploit to new habitats, and to quickly become established in new locations, represents a substantial public health challenge [66, 67].
Dengue virus is a key example of an arbovirus that has co‐evolved and spread with its vectors to primarily infect humans as hosts, and in urban rather than rural habitats [20, 68]. Dengue is a Flavivirus that has been spreading around the globe for centuries [69]. The first well‐documented epidemics occurred during the 18th century, though were sporadic and not considered a major public health problem [20, 69, 70]. This
Chapter 2: Literature review & Research objectives 24
changed with the movement of troops around the world during World War II (WWII), which spread both viruses and their Aedes vectors on an unprecedented scale, instigating the current pandemic [34]. Dengue is now the most prevalent arboviral disease in the world, endemic in >100 countries, and causes as estimated 390 million infections every year [18, 19].
Other important arboviruses in the same family as DENV include WNV, ZIKV and YFV. WNV is the most prevalent arbovirus globally, after DENV, having been effectively spread via a wide range of migratory avian host species, as well as a number of different mosquito vectors, primarily Culex (Cx.) spp. [10]. Before 1996, WNV was not considered highly pathogenic. However, the emergence of a more virulent genotype preceded introduction of the virus into North America, where WNV has become a major public health problem [71]. Up to 80% of infections are asymptomatic; however, severe neurological disease that occurs in <1% of cases can be fatal [72]. Transmission risk is associated with human‐modified environments including urbanised and agricultural areas, where vectors and hosts are abundant [73].
WNV, ZIKV and YFV originated from forests in Africa, where they existed in sylvatic cycles between non‐human primates and arboreal mosquitoes [23]. Like WNW, ZIKV was initially considered a relatively innocuous pathogen with sporadic human infections for 50 years prior to emerging in the Pacific and the Americas [74]. It is now been associated with Guillain Barré syndrome and congenital malformations (microcephaly) in Oceania, the Americas, Asia and Africa [10, 75]. YFV has caused regular outbreaks and morbidity in African nations and in Central and South America since at least the 17th Century, with a fatality rate of between 20‐50% [26, 76]. It can exist is both sylvatic or urban cycles, and continues to cause major epidemics as well as importations in returning travellers, as occurred in China in 2016 [27] and 5 different European countries in 2018 [77].
Similarly, human travel and trade over the past 2 centuries enabled CHIKV to spread and adapt to new environments and vectors [10, 23, 78]. CHIKV is an alphavirus, part of the Togaviridae family, but shares epidemiological and ecological features with the flaviviruses DENV, YFV and ZIKV [62, 79]. CHIKV infection causes severe, polyarthralgic disease, an increased risk of Guillain Barré syndrome, and can contribute
Chapter 2: Literature review & Research objectives 25 to fatal complications in a small proportion of elderly cases where there are co‐ morbidities [80]. Its similarity in disease symptoms with DENV and Zika means that in areas where all three circulate, it can often be misdiagnosed. CHIKV emerged from its native sylvatic foci in Sub‐Saharan Africa and spread into urban transmission cycles around the globe, most recently arriving in the Americas in 2013 [62].
Emergence and spread of each of these epidemic arboviruses have been associated with a) introduction of the virus into large, naive host populations, b) presence of receptive (immunologically naïve) human and/or wildlife hosts, and c) introduction or adaptation of competent vector species [81, 82]. Many emerging arboviruses are spread by common vectors; in particular the Aedes species: Ae. aegypti and Ae. albopictus, which have played a major role in the emergence of the global pandemics caused by CHIKV, ZIKV and YFV. However, Culex species are important in the transmission of arboviruses such as WNV and Japanese Encephalitis virus (JEV) [30], and in some cases additional mosquito species can act as ‘bridge’ vectors to transport viruses between sylvatic and urban habitats [81, 83‐86].
Selective pressures on viral evolution also played a role, with environmental change stimulating viruses to adapt to new vectors and hosts [4, 6, 8‐10]. Urban expansion and economic development activities that disturb or encroach upon natural ecosystems can facilitate arboviral introduction to new vector and host species [66]. For example, changes in land use, such as conversion of forests to agricultural or urban uses, can significantly alter the wildlife species diversity present, as well as vector‐host dynamics [67, 87, 88]. This may either increase (amplification effect) or decrease (dilution effect) circulation of arboviruses, or pathogen exchange between habitat types, depending on the species balance present [30, 89, 90]. However, disease emergence is a multifactorial process – with evolutionary pressures operating at different levels on pathogens, hosts and ecosystems [91]. Thus, specific causal events that link environmental change to risk of arboviral diseases are often context‐specific.
Chapter 2: Literature review & Research objectives 26
Dengue epidemiology Dengue disease is caused by one of four Flavivirus serotypes (DENV1 – DENV4) that each evolved from a sylvatic ancestor between 140‐350 years ago [20, 92]. Infection causes a febrile illness with a wide spectrum of symptoms, ranging from asymptomatic or mild illness, to severe and potentially fatal haemorrhagic disease [93]. Symptoms occur following a viral incubation period of 4‐8 days following transmission by an infected mosquito, and may include headache, retro‐orbital pain, body aches, nausea and vomiting, joint pains, weakness, and rash, alongside regular spikes in fever, and in some cases may progress to include potentially fatal internal haemorrhage [93, 94]. Given that the primary symptoms may present similarly to those of a number of other infectious agents (e.g. leptospirosis, rickettsial diseases), definitive diagnosis requires diagnostic testing by either viral RNA or protein detection, or via serology [95, 96]. While the available tests are both sensitive and specific for dengue, they may not always be available in resource‐limited settings. Hence, in many endemic countries, diagnosis still relies largely on clinical observations.
The first references to a dengue‐like illness were in China more than 1000 years ago [69, 97]. The first reports of large epidemics were associated with the slave trade, and the movement of military personnel throughout the 17th and 18th centuries, and were reported primarily from port cities and towns [97]. Larger epidemics appeared more widely throughout the Americas from 1635 (Martinique, Florida, Guadeloupe, Mexico, Panama, and Philadelphia [70]. Reports in East Africa and Asia commenced from the late 19th century until the early 20th century, from countries including Zanzibar (1870), Tanzania (1871), Singapore (1875), Yemen, India, Pakistan and Egypt (1887), Burma (1888), the Philippines, Thailand and in Malaysia from 1901, but were still relatively sporadic [98, 99].
The next major wave of epidemics occurred during and after the Second World War as a result of dramatic increases in human movement, and the creation of favourable environmental and socio‐economic conditions to support transmission [69]. It was at this time that the virus was first isolated: first from an outbreak in 1943 amongst returning solders in Japan (Serotype 1), followed by identification of a second serotype in Hawaii in 1945, and the fourth and fifth serotypes during outbreaks in the
Chapter 2: Literature review & Research objectives 27 Philippines and Thailand in 1954 [100‐102]. It was shortly after WWII that the severe (haemorrhagic) form of disease was recognised in South East Asia and that many countries in Asia became hyperendemic for all 4 viral serotypes [103]. This coincided with a surge in economic growth and rapid urbanisation in many South East Asian countries, along with the presence of large populations which could support frequent, high‐magnitude epidemics [33].
From there, the virus spread to South Asia and the Americas, and increasing incidence and co‐circulation of viral serotypes have been reported in most regions of the world [20, 70]. Although the development and accessibility of diagnostic testing, and creation of surveillance programs has contributed to this rise, the primary drivers of dengue disease expansion are the spread and adaptation of dengue vectors to highly‐ populated urban environments in several tropical countries [14, 17]. Additionally, circulation of multiple DENV serotypes, or a switch from one serotype to another, is associated with increased risk of the severe (haemorrhagic) form of dengue [93, 104, 105]. This is due to the immunological effects of viral infection, where the antibody response to infection with one serotype can potentially enhance that of subsequent infections [106].
Dengue transmission occurs as a result of interactions between human or wildlife hosts, mosquito vectors and viruses in the presence of suitable environmental conditions [94]. Mosquito vector dynamics are a key determinant of dengue transmission, providing the vehicle for the virus to spread and survive within various hosts and environments. Although there are several vectors that can be involved in the transmission of DENV, international spread of the principal DENV vector, Ae. aegypti, has been the major driver of DENV’s global geographic expansion [99]. However, both Ae. aegypti and Ae. albopictus have been implicated in epidemics, and they may either co‐breed, or displace each other in terms of dominance of a particular habitat [43, 107‐ 109]. Following DENV infection, mosquito vectors remain infected for life, and are also capable of vertical transmission of the virus to their offspring [110]. This allows individual mosquitoes to continually transmit DENV to multiple hosts.
Many cities and rural towns in South East Asia now contain a mixture of both species, surviving within overlapping ecological niches and host range [111]. A key
Chapter 2: Literature review & Research objectives 28
challenge for the control of dengue is the resilient, opportunistic, and highly adaptive nature of these two primary mosquito vectors. Both species have expanded their geographical range within the last 50 years, and this is predicted to continue in the coming decades, in response to the continued urban development and the favourable conditions potentially created by global warming [33, 90, 112, 113]. While Ae. aegypti is highly anthropophilic in its biting behaviours, Ae. albopictus feeds on a broader host range including both humans and animals – and it likely mediates the overlapping transmission between human and sylvatic cycles [85, 86].
DENV reservoir hosts may be either human or animal, depending on their availability and the biting preference of vector species [114] (Figure 1). An enzootic forest cycle of DENV transmission occurs in the forests of Malaysia, Thailand and Senegal between vectors Ae. niveus spp. and Ae. albopictus and primates in the genera Macaca and Presbytis [38, 85]. The circulation of sylvatic DENV variants in humans, and their relative role in amplifying them, is not clear. Future changes in climate, land use and/or urbanisation processes could have potential to expand the distribution of sylvatic DENV genotypes beyond forested environments – by changing the distribution of forest‐ dwelling vectors and hosts [38].
Climate has been proposed as an important driver of the occurrence and spread of dengue [115, 116], with several climate‐based models developed investigate its influence as a disease driver, to predict outbreak conditions, and forecast vector geographic expansion [117‐121]. While most studies agree that climatic factors have some influence on vector breeding, others have found no specific correlation between climate and outbreaks [122‐124]. This could be because the impact of climate on disease drivers acts at different levels, and these are not always easily measured. For example, the local micro‐climate of either natural or artificial vector habitats could have greater influence on disease than broader‐level (e.g. city‐wide, or global) climate variables [125, 126]. Additional influences on virus transmission probably interlink with climate variables, such as local socioecological dynamics, patterns of human movement, and vector control measures being implemented; hence, dissecting the specific influence of climate alone can be challenging [127‐130].
Chapter 2: Literature review & Research objectives 29
Figure 2.1. Transmission cycle of dengue.
DENV circulates through human and primate hosts, and within several mosquito species (primarily Ae. aegypti and Ae. albopictus), in both urban and rural environments.
The recent resurgence of dengue appears to be closely associated with demographic and societal changes over the past 50 years [34]. Two of the major driving forces are unprecedented global population growth, and the associated uncontrolled urbanization. Population growth, and rapid influx of people from rural to urban centres, often results in sub‐standard housing, crowding, and poor water, sewer, and waste management systems [33]. These create ideal conditions for increased transmission of mosquito‐borne diseases, particularly in tropical urban centres of developing countries [131, 132]. A third major factor has been the lack of effective mosquito control in areas where dengue is endemic [32, 133]. Other equally important key factors already mentioned are movement of people and viraemia, significant ecological changes occurring in endemic regions, and invasion and establishment of vectors.
Ross River virus epidemiology
RRV is an alphavirus that belongs to the Togaviridae family, closely related to Chikungunya, Barmah Forest and Sindbis viruses [134, 135]. Since its initial isolation from a pool of Ae. vigilax mosquitoes in 1963 [136], four distinct genetic variants have
Chapter 2: Literature review & Research objectives 30
been identified: GI to G4, with each type historically thought to predominate in different regions of Australia or the Pacific Islands [137‐140]. While there is evidence of geographic separation of genotype sub‐lineages, substantial co‐circulation of different regional genotypes also occurs. [139, 141]. These four genotypes are estimated to have diverged from a common ancestor in the early 20th century, with periodic bursts in genetic diversity occurring roughly every decade [141]. Emergent lineages have been rapidly geographically dispersed within a relatively short time frame [141], which may suggest spread by wide‐ranging host species, such as humans and/or birds. However, the biological and ecological drivers of RRV genotypic isolation or dispersal are not known.
Infection with RRV can cause a polyarthritic syndrome in humans that can persist for months after initial infection [135, 142‐144]. Clinical symptoms can include fever, headache, rash, lethargy and joint pain, which can make it difficult to distinguish from other arboviruses [145]. Hence, diagnosis requires confirmation of the aetiological agent, most commonly by serology [146]. The disease consistently affects both male and female adults between 30 and 50 years old, rather than children [49, 54, 147]. A high proportion of infections may be asymptomatic, undiagnosed or misdiagnosed due to inconsistent symptoms [135, 148]. Although the clinical course of disease is fairly well understood, no specific treatment is available for RRV disease [54, 145]. A vaccine candidate has undergone clinical testing; however, its commercial viability is uncertain [149‐151].
The disease caused by RRV infection was first described during an epidemic of polyarthritis in Southern New South Wales in 1928 [152], although it’s possible epidemics occurred throughout earlier periods of Australian settlement [147, 153, 154]. The virus was first identified from an Aedes vigilax mosquito in the Ross River region of North Queensland in 1958, and was first isolated from a patient suffering from epidemic polyarthritis in 1979‐80, during the largest recorded outbreak in the South Pacific islands [155‐157]. More than 50,000 cases were recorded during this major epidemic, which occurred across several Pacific Island countries, including American Samoa, Fiji, Cook Islands and New Caledonia [157]. This epidemic was thought to have originated from a viremic Australian traveller to Fiji, and to have spread via human‐mosquito‐human
Chapter 2: Literature review & Research objectives 31 transmission, possibly with the contribution of domestic animals as short‐term amplifying hosts [56]. Since then, no major epidemics have occurred in these island countries, and it was assumed that no ongoing transmission was present after 1980. This assumption was based partly on the lack of additional outbreaks in the Pacific Island countries, but also on the prevailing thought that marsupials (endemic only to Australia and PNG) were essential to the RRV transmission cycle.
RRV circulation has been documented in PNG and the Solomon Islands since 1975 [158, 159]; however, the lack of routine surveillance or research studies means actual distribution and prevalence are unknown [159‐161]. From the early 1990’s, RRV cases were sporadically reported in tourists who had visited Fiji [162, 163]. More recent evidence from the Pacific Islands indicates that RRV may have continued to circulate undetected in countries outside of Australia for decades. Seroprevalence studies conducted in American Samoa and French Polynesia from 2010‐2013 found unexpectedly high RRV antibody prevalence (74% and 34%, respectively), with the majority of tested samples from residents with no travel history [55, 56]. This was the first detection of RRV circulation in French Polynesia, and the first evidence of endemic RRV circulation in American Samoa after the only known RRV outbreak in 1979‐80 (seroprevalence in residents born after 1980 was 63%) [56]. Follow up studies in Fiji and French Polynesia between 2013 and 2015 confirmed that ongoing endemic circulation of RRV also occurs there [57, 58].
The highest prevalence and most frequent epidemics of RRV are reported from Australia. Approximately 5,000 cases are notified each year, of which a persistently high proportion from Queensland [46]. Notifications typically peak annually between January and April although seasonal trends vary between states and territories [46, 148]. The largest Australian epidemic (9,550 cases) occurred in 2015, with the majority of cases (6,193) recorded in Queensland [164]. Subsequently, following significant rainfall events in early 2017, the states of New South Wales and Victoria reported higher than average numbers of mosquitoes, as well as a marked increase in RRV case numbers [46, 165]. For Victoria, it was the largest number of annual cases ever recorded, and the first record of cases in outer metropolitan areas of Melbourne city [166, 167]. An increase in cases was also recorded in several areas of Western Australia in March, 2017, with nearly
Chapter 2: Literature review & Research objectives 32
triple the usual number of cases in Perth city, prompting public health warnings [46, 168].
Ross River virus is a vector and host generalist. It has been isolated from ≥40 species of freshwater and estuarine mosquito vectors, as well as several vertebrate host species, over a wide geographic range [49, 53, 169] [Figure 2.2]. Marsupials are considered a major RRV reservoir, though it is likely that the virus is maintained by additional hosts [45, 53]. This diversity of RRV host species likely supports long term persistence of the virus [139, 170, 171], while also imposing selective pressure on its genetic diversity [141]. The specific vectors and hosts responsible for both maintaining the natural cycle of RRV and for mediating human epidemics are unknown, and studies to differentiate these key drivers between habitat types are lacking. Existing hypotheses about important vector and host species involved in RRV circulation are derived from sporadic viral isolation and experimental infection studies; however, the relative role of various species as amplifiers of the virus is unclear [45, 49, 53, 54, 171].
The evidence for competent mosquito vectors is relatively strong, compared to that for reservoir hosts. Of >40 mosquito species from which RRV has been isolated, >10 have been implicated as competent RRV vectors in laboratory studies [49]. Additional species possibly also contribute to the maintenance of RRV, which have not yet been confirmed by empirical studies. Nonetheless, it is clear that a range of vector species are capable of maintaining RRV, and these are present across diverse habitats and bioclimatic regions, both in Australia and internationally [172‐176]. The Australian species most commonly associated with epidemics are: Ae. vigilax and Ae. camptorhynchus in rural salt marsh and coastal regions, Cx. annulirostris and Coquillettidia linealis in inland regions; and Ae. notoscriptus in peri‐urban and urban areas [177‐181]. Some vectors are geographically restricted to specific bioregions of Australia; thus, vectors important for maintaining transmission may differ between regions. For example, the vector Ae. camptorhynchus is not present in northern Australia, but maybe responsible for significant transmission in South Australia [182, 183]. Conversely, in northern Australia, the saltmarsh species Ae. vigilax has often been implicated though it is not present in southern Australia [177, 181, 184]. Other less‐ reported species, such as Ae. sagax and Verrallina (Ve.) sp., have also been associated
Chapter 2: Literature review & Research objectives 33 with outbreaks [54, 185, 186]. The relative importance of these vectors during epidemics is not known, though the density and distribution of RRV vectors, in tandem with the presence of potential reservoir hosts within a given area, provides clues regarding possible drivers of disease transmission [171, 187, 188].
Figure 2.2. Transmission cycle of RRV. RRV circulation is maintained by multiple marsupial, mammal and bird species, as well as >40 mosquito species. The virus is maintained across both urban and rural environments, but the ecological determinants of its distribution and spread are unknown.
In Queensland, large outbreaks in Brisbane and the Sunshine Coast Region during the 1990s were associated with high abundance and virus isolation from freshwater vectors Cx. annulirostris, Ae. notoscriptus, Ae. procax, and Ae. vittiger as well as saltwater vectors Ae. vigilax, Cx. sitiens and Ve. funerea, with differences noted between coastal versus inland areas [177, 187, 189, 190]. The most recent outbreaks of 2014‐ 2015 were associated with increased abundance of the freshwater mosquito species Cx. annulirostris and Ae. procax in Brisbane following high rainfall [51]. However, abundance alone is likely not an adequate indicator on its own to incriminate vectors, as not all abundant vectors will be important in human spillover. Given that a number of mosquito species within a given location may all be competent to transmit RRV, and that the host‐ feeding behaviours of many of these overlap, it is possible that different species
Chapter 2: Literature review & Research objectives 34
contribute to epidemics simultaneously, or at different times during an outbreak [177, 191, 192]. To definitively determine this, vector incrimination studies, including viral isolation from vectors in association with human outbreaks, are much needed. Unfortunately, these have rarely been performed, and in their absence, vector abundance is often used as a proxy.
The potential circulation of RRV outside of Australia and PNG also suggest the need to reconsider common assumptions about amplifying hosts of RRV. While there is general agreement in the literature that marsupials are the primary reservoirs of RRV [45, 47, 49, 169], evidence from Pacific island studies indicate that RRV persistence is possible within alternative hosts. In Pacific Island countries, where marsupial hosts are absent, the species with potential to act as amplifying hosts may include rodents, pigs, cows, horses, dogs, cats, birds and bats; however, few studies have investigated these [193‐197]. It is also possible that RRV could be maintained in purely human‐mosquito‐ human cycles. This has been suggested to have occurred in the Pacific Island outbreaks of the late 1970s, and by increases in urban‐associated outbreaks in Australia, although empirical evidence is unclear on whether human infection alone is sufficient to sustain epidemics [49, 171].
A 2017 modelling study considered the key attributes of hosts required to contribute to RRV epidemics in different transmission contexts (rural/urban, enzootic/epizootic/epidemic), and found that the most likely species with sufficient host competence and abundance to contribute to RRV epidemics are humans and possums. [198]. There is currently insufficient evidence for host competence, and this limits any model generated. But the available information suggests that although marsupials are competent reservoirs of RRV, there is also potential for placental mammals and birds to regularly act as amplifying hosts of the virus [53, 191]. Additionally, host species contributing to epidemics may be different to those that maintain endemic circulation of RRV. Given the limited number of host studies to date, future studies would benefit from a broader investigation of potential hosts involved in RRV transmission, especially given the diverse environments in which RRV circulates.
In Australia, RRV is transmitted in both coastal and inland regions, and across a range of climate zones. The virus is maintained within a range of habitats including
Chapter 2: Literature review & Research objectives 35 freshwater and estuarine wetlands, coastal regions, salt marshes, floodwaters, established wetlands and urban areas [49, 54, 199]. Several studies around Australia have demonstrated the highly varied contexts of RRV transmission [147, 169, 200, 201]. There is a broad range of factors that can influence RRV transmission on long‐ and short‐ term scales including human behaviours, host population immunity, virus‐host adaptation, climate or weather patterns, proximity to mosquito habitat, mosquito vector densities and survival, and the transmission rate between vectors and hosts. The temporal and spatial dynamics of these parameters have been suggested to explain the changing patterns of RRV cases observed across Australia during recent decades [52, 199, 202‐204], but the tremendous dynamism of these patterns presents a challenge to identifying and controlling the major mechanisms of virus transmission.
Previous studies of RRV outbreaks describe a heterogeneous spatial pattern of disease, with variation according to different geo‐climatic regions of Australia [154, 200, 205, 206]. Various models have been constructed to investigate the variables linked to RRV disease distributions, primarily assessing the role of climate, vector and host prevalence, and socio‐ecological factors [154, 169, 170, 201, 207‐211]. The results have varied according to the ecological region and the combination of transmission factors studied, with many authors concluding that the complexity in RRV transmission dynamics requires more detailed assessments of the local socio‐ecological context [154, 169, 206, 207, 209, 212, 213]. Analyses to date have also emphasized the need to clarify the relative risks of vector and host distribution for human infection [170, 204, 209, 214‐ 216].
In particular, analysis of the broad range of infection risk factors such as proximity to mosquito and vertebrate host habitats, correlation with climatic conditions, and vegetation cover including bushland and wetlands is required at a local level [204, 209, 212, 215, 216]. The linkages between these risk factors and arboviral circulation in peri‐ urban and urban areas are not sufficiently understood. Greater understanding of the links between climate factors such as temperature, relative humidity, rainfall and high tides, and RRV disease risk is also needed. These factors directly impact the reproductive capacity and survival of freshwater and estuarine mosquito vectors of RRV, and could partly influence the periodic availability of vectors and hosts, if abundance and
Chapter 2: Literature review & Research objectives 36
reproductive cycles are linked to weather suitability [188, 217‐219]. Climate alone does not predict RRV outbreak occurrence, but it is a requisite precursor to outbreaks [169, 206, 220]. As with dengue, the relative importance of climate and other socioecologic drivers of infection (such as the availability and proximity of competent vectors and susceptible hosts) is difficult to separate, as their influences on outbreak occurrence are likely interlinked.
Prevention and control of arboviral disease
The potential for arboviral expansion and emergence is a global public health concern and provides impetus to develop the knowledge and tools to prevent future human health and economic burden. With the exception of yellow fever, few arboviral diseases have a commercially viable, effective vaccine, nor medications that specifically treat human infections [135]. This includes both DENV and RRV, although attempts to develop vaccines for both diseases have been made [150, 221]. Prevention efforts thus focus on vector control, as the most effective intervention currently available, although the choice of tools that are safe, efficient, cost‐effective and that suit the full range of vector habitats and human behaviours are limited [2, 32, 222].
For arboviruses, especially dengue, urban areas are a key focus of control efforts. In particular, the presence of artificial breeding containers is a major problem [223, 224]. Methods to control vector populations include larviciding of natural and artificial water storage sites, and prevention of water accumulation through environmental modification and waste disposal [2]. Reducing adult vector populations through insecticide spraying is a key strategy during outbreak conditions; however, may not be an ideal long‐term control strategy [225, 226]. Alternative strategies, such as community‐based environmental management approaches that engage communities to treat, cover or empty water‐holding containers, shows promise as a sustainable and cost‐effective approach [225, 227]. These might be complemented by human‐targeted approaches which aim to reduce the risk of human contact with vectors. These may include use of window screens in households, use of protective clothing and mosquito repellents, and modification of human behaviours [228‐230].
Chapter 2: Literature review & Research objectives 37 For dengue vector control, the current strategy advocated by the WHO is the Integrated Vector Management (IVM) approach, which aims to integrate community‐ based environmental management with epidemiological and entomological surveillance and mosquito control [231‐233]. The objective is to achieve dengue control in a more cost‐effective and sustainable manner, engaging the efforts of the local community to encourage social responsibility towards reducing mosquito breeding and biting. Vector control programmes and communities in endemic countries also carry out active monitoring and surveillance of human and vector populations to aid prevention and response efforts, as well as determining effectiveness of existing interventions [234‐ 236]. However, while integrated approaches have resulted in reductions in breeding sites and entomological indices, it is not yet clear how well this translates to an impact on disease prevention and health outcomes [237]. Current control strategies have often not been effective in preventing epidemics or arbovirus spread, and thus there is a need to investigate new tools, and to further optimise existing approaches [227]. Specific challenges to combat include effectively targeting cryptic domestic mosquito breeding sites, addressing the development of insecticide resistance in vectors, and development of strategies to building capacity of communities to implement vector control. [238]
For RRV, vector control in Australia aims to reduce abundance of multiple species in both salt and freshwater habitats. Local governments are responsible for vector control activities, and they invest significant effort toward larval control of nuisance mosquitoes associated with mangrove swamp and saltmarsh environments adjacent to residential areas, through aerial insecticide spraying. Ground larviciding of freshwater larval habitats is also performed as needed, and as dictated by available local government resources. These activities aim to reduce overall abundance of a range of RRV vectors in different habitats, without necessarily targeting specific RRV vectors. In the absence of identification of specific vectors that are linked to RRV outbreaks, a generalised approach to reduce mosquito abundance in proximity to human habitats is the next best available option to control RRV [239].
Local government costs for vector surveillance and control have been reported to be several million dollars per year [239, 240], while individual treatment seeking and health system costs for Ross River virus disease are estimated to cost at least US $4
Chapter 2: Literature review & Research objectives 38
million/year [45, 206, 241, 242]. Given these costs, it is perhaps surprising that Australia does not have targeted prevention strategies for RRV. More targeted strategies are hindered by the absence of definitive evidence incriminating specific vectors and hosts in human epidemics [45]. Aside from vector control activities to reduce overall mosquito abundance, management strategies include compulsory notification of arboviral disease diagnoses, promotion of personal protective measures, as well as use of molecular viral surveillance strategies [47]. Case definitions have also been revised to increase diagnostic accuracy [148].
To more effectively prevent RRV transmission, a combination of measures is required to interrupt vector breeding and biting, and to reduce viral circulation in hosts. This relies on an adequate understanding of the relationship between specific vectors, the amplification dynamics of the virus, and where and when the highest risk to human populations occur [243]. This might be enabled though enhancing viral surveillance in vector species, and clarifying the contribution of additional risk factors (such as human movement and behaviour, landscape factors, vertebrate host types present in high risk locations) for virus spillover to humans [244]. This knowledge could enable design of more efficient and cost‐effective targeting of vector control to specific mosquito species, alternative approaches to reduce RRV exposure, and ultimately fewer RRV infections.
Chapter 2: Literature review & Research objectives 39 Summary and implications The increasing incidence of both dengue and RRV disease, and potential adaptation of disease vectors and hosts to new locations, emphasises the need to investigate new approaches to prevent future epidemics. However, the complexity and variability of the epidemiology and control of both diseases has made the development of effective preventative approaches challenging. Improved approaches for management of both viruses require clearer quantification of the drivers of disease, including the relative contribution of specific mosquito vectors, and greater understanding of how vector‐host interactions in diverse and changing environments are linked to disease incidence. Some of the key characteristics and risk factors relevant to consider in the control of each virus are summarised in Table 2.1.
For dengue, there is an urgent need to design effective and sustainable vector control strategies to reduce the substantial human burden of this disease, and limit further expansion of the virus. This requires an evidence‐based approach that is informed by knowledge of the virus’ ecological dynamics, and how and where interventions could be best targeted. Particularly in urban areas of South East Asia, where the disease burden is highest, improved understanding of the disease ecology could help educate endemic communities, and develop more effective tools to reduce Aedes breeding. Improved targeting of interventions, in combination with enhanced surveillance of viral serotypes in humans and in vectors, could help limit the persistent threat of this virus to public health.
For RRV, there is also rising public health impact the disease in Australia, and a risk of expansion to affect additional countries. Despite this, the specific risk factors determining virus spillover to humans have not yet been identified. Major gaps in knowledge include identification of the specific mosquito and vertebrate host species responsible for mediating human epidemics of RRV, and understanding of how these differ across environments.
Chapter 2: Literature review & Research objectives 40
Table 2.1. Comparison of epidemiological characteristics of dengue and RRV.
Characteristic Dengue Ross River virus RNA virus of the Flavivirus RNA virus of the Alphavirus Virology genus, with 4 serotypes. genus, with 4 genotypes. Global (currently 129 Australia and Pacific Island Geographic distribution countries). Countries and Territories. Several vectors across Two primary vectors: Ae. multiple genera; thought to aegypti and Ae. albopictus Vector species be primarily Aedes and associated with freshwater Culex spp. Both freshwater habitats. and estuarine habitats. Thought to include humans, Host species Humans and primates. horses, marsupials, and other unknown species. Urban residential areas, Peri‐urban residential Environmental associations rural towns. areas, rural towns. Varies in each country; often children under 15 In Australia, primarily adults Demographic groups at risk years, but adults aged 30‐60. increasingly affected. Fever, headache, muscle and joint pain, skin rash. Common symptoms of Fever, headache, muscle Severe haemorrhagic infection and joint pain, skin rash. complications can lead to dengue shock syndrome. Combination of clinical features and either Combination of clinical detection of either viral features and Diagnostic method antigen, IgM or IgG seroconversion from IgM to antibody and/or molecular IgG antibody. detection of virus. No specific treatment. Hydration, analgesic and No specific treatment. anti‐inflammatory drugs; Analgesic and anti‐ Treatment intravenous fluid and inflammatory drugs for possibly intensive care if symptom relief. severe. Recovery or death. Risk of Recovery, with possibility of subsequent severe disease Disease outcomes longer‐term fatigue, muscle upon reinfection with a and joint pain. different serotype. General vector surveillance Aedes surveillance and and control. Vaccine control. Vaccines being Prevention and control available but has not tested, but not yet progressed to commercial available. development.
Chapter 2: Literature review & Research objectives 41 Factors that facilitate RRV circulation within urban environments, or between rural and urban environments, are not well‐studied. Research to date has provided snippets of information on the ecological drivers of RRV transmission, and while there has been a great deal of conjecture around the specific drivers of outbreaks in different parts of Australia, there is a paucity of empirical investigations or proof. This limited knowledge of the specific determinants of RRV infection, particularly which vectors and host mediate transmission in different environments, hinders the development of more targeted prevention and control strategies. A more detailed understanding of the ecological and human factors leading to spillover and maintenance of RRV disease could inform future public health decision making and preventive strategies, both for RRV and potentially other arbovirus infections.
Although dengue and RRV differ in geographic distribution, relative disease burden, and have vastly different impacts in terms of morbidity and mortality, they share some common gaps in knowledge required to improve their prevention and control. Prevention methods for both viruses rely entirely on vector control, which for both diseases is currently limited in its capacity to prevent vector breeding and biting. Despite being transmitted by quite different vector species, the inability to rapidly detect and effectively respond to the presence of infected vectors – across the range of endemic environments – is a shared challenge for both viruses. Effective surveillance and response to both viruses is currently impeded by the resource requirements of vector surveillance, and the unavailability of control tools that are both targeted and can be applied with sufficient coverage to effectively interrupt transmission. Both also lack specific treatments, and have substantial health and socioeconomic impacts that are likely to persist well into the future.
Given the continuing public health threat posed by dengue and RRV, and likelihood of new arboviral disease emergence, it is essential to enhance current capacity to respond to future epidemics. This might include further development of vaccines or treatments to reduce health impacts. However, there is a broader need to address the underlying drivers of arboviral emergence and ongoing circulation. These drivers include climate, environmental change, urban development and globalisation which may act at both local and global levels, and across disciplines. In terms of practical and immediate
Chapter 2: Literature review & Research objectives 42
actions, a core approach to better combat arboviruses is to improve effectiveness of vector control, and this requires enhanced surveillance methods, and a greater understanding of virus‐vector‐host dynamics.
Research objectives The purpose of this research was to identify and test potential epidemiological determinants of rural and urban transmission for two arboviruses, dengue and Ross River viruses. It employed two major approaches: firstly, the use of spatial epidemiological techniques to retrospectively assess disease patterns of dengue and RRV, and to infer transmission risk factors; and secondly, the conduct of prospective field and laboratory studies designed to investigate local RRV transmission dynamics in more detail. This research was guided by four objectives (listed below): the first two relate to the planned retrospective epidemiological analyses, while the third and fourth objectives focus on the prospective collection and analysis of field data.
Specific objectives of this research were to:
1) Explore spatial and temporal patterns of dengue disease across urban and rural environments of Sabah, Malaysia, through analysing retrospective data of dengue notifications and routine vector surveillance at the finest available geographical scale;
2) Investigate potential drivers of RRV disease patterns across urban and peri‐urban environments of South East Queensland, Australia through fine‐scale spatiotemporal analysis of RRV disease patterns in relation to urban and rural locations, and landscape factors;
3) Investigate novel techniques to aid the investigation of RRV transmission dynamics, specifically to maximise the information that can be inferred from mosquito blood‐meal analyses; and
4) Investigate relationships between vector‐host interactions and RRV transmission in urban and suburban locations through the assessment of mosquito feeding behaviours, relative to available host species.
Chapter 2: Literature review & Research objectives 43 Through these objectives, this research broadly sought to extend upon the existing understanding of arboviral disease transmission across different endemic settings, with a view to providing a platform for future research, and ultimately for informing future disease control and prevention strategies. A secondary aim was to provide a framework for investigation of other vector‐borne diseases, in consideration of their rising global public health importance.
Significance and Scope
For the two viruses studied, I used spatial and temporal data analysis methods and descriptive observations of disease trends to produce up‐to‐date reports of trends for each disease, across urban and rural environments. For both diseases, my investigations were guided by availability of public health notification data, as well as of complementary data sets including population (census) data, and geospatial data detailing geographic boundaries of the locations of interest. For DENV, access to detailed data sources for retrospective analyses was more limited than for RRV. The analyses were based on a public health dataset of dengue notifications, which presented some limitations in terms of variables and spatial scale available to analyse; however, it was temporally detailed, and provided rare access to entomological surveillance data collected as part of outbreak monitoring. It was also unique in that although dengue is highly endemic across Malaysia, epidemiological investigations have been neglected in rural areas, and trends have not been previously examined for the state of Sabah. The work in this thesis helps quantify the burden of this important arboviral disease and describes potential links to its underlying determinants.
A more in‐depth assessment of epidemiological trends was possible for RRV than for DENV due to data of increased breadth and detail being available. Previous studies of RRV epidemiology in Queensland have often encompassed a narrow selection of risk factors, at‐risk locations and time frames. I assessed RRV trends in finer spatial and temporal detail than previously performed in the South East Queensland (SEQ) region, and provided the first longitudinal analysis (over 16 years) of RRV disease trends, including assessment of demographic and environmental trends. Unfortunately, I was not able to include data for the full range of ecological correlates of infection in the
Chapter 2: Literature review & Research objectives 44
retrospective analyses, such as entomological data, land cover data and climate data. These would be valuable to investigate within a multivariate model, but this was outside the scope of this PhD.
Because RRV is much more understudied than DENV, I chose to focus on RRV to further explore its transmission dynamics through field and laboratory investigations. This involved an empirical study employing both ecological field surveys, mosquito sampling and the use of a new laboratory xenodiagnostic technique – and extended upon previous studies which have primarily investigated only a single component of transmission. The field study was not designed to enable a full assessment of RRV transmission pathways – rather, it was planned as a pilot study to test a novel investigative approach; yet, it was able to shed light on transmission dynamics in Brisbane, and provide a stimulus for future investigations. The methodology developed for this component of my research also has relevance for the study of broader vector‐ borne and zoonotic diseases.
Although the relative attention given to each virus in this thesis is unequal, it nevertheless contributes to the literature on both DENV and RRV ecology in high‐burden regions. It also extends existing knowledge of potential transmission risk factors for both viruses in the context of environmental change and increasing urban expansion. Investigation of both dengue and RRV highlighted the unpredictability of arboviral disease in terms of the potential for future outbreaks, especially in the presence of continued global change and its associated impacts.
Chapter 2: Literature review & Research objectives 45 Chapter 3: Results for Objective 1
This chapter includes the following manuscript:
Murphy AK, Rajahram GS, Jilip J, Maluda M, William T, Hu W, Reid S, Devine GJ, Frentiu FD. Incidence and epidemiologic features of dengue in Sabah, Malaysia.
This manuscript in Chapter 3.1 describes spatial and temporal trends of dengue notifications for the state of Sabah, Malaysia. It uses a public health data set from a rural state of Malaysia, and applied spatial and temporal analysis methods to report epidemiological trends for state‐notified dengue cases over a 7‐year period. The findings of this chapter contribute to understanding of an important arboviral disease in an understudied area of South East Asia, and highlight the importance of assessing disease risks across a variety of ecological settings and at‐risk populations.
This chapter begins the exploration of how and why arboviral transmission patterns might vary across rural and urban environments, and was important in the development of research techniques to be applied in Chapter 4. The inclusion of vector surveillance data in the analyses was unique in that it extended upon the use of purely demographic or clinical variables often employed in epidemiology studies. This enabled a broader investigation of possible entomological and environmental contributors to the patterns observed.
Additional analyses to detect space‐time clusters of cases were also performed on this data set, but were not included in the final manuscript. These are shown in Section 3.2, as they were important for refining spatiotemporal analysis techniques, and for exploring the use of appropriate spatial and temporal scales in assessing disease patterns.
Chapter 3: Results for Objective 1 46 Statement of Contribution of Co-Authors for Thesis by Published Paper
The Co-authors Rajahram GS, Jilip J, Maluda M, William T, Hu W, Reid S, Devine GJ, and Frentiu FD of the manuscript below have certified that:
1. they meet the criteria for authorship in that they have participated in the conception, execution, or interpretation, of at least that part of the publication in their field of expertise;
2. they take public responsibility for their part of the publication, except for the responsible author who accepts overall responsibility for the publication;
3. there are no other authors of the publication according to these criteria;
4. potential conflicts of interest have been disclosed to (a) granting bodies, (b) the editor or publisher of journals or other publications, and (c) the head of the responsible academic unit, and
5. they agree to the use of the publication in the student's thesis and its publication on the QUT's ePrints site consistent with any limitations set by publisher requirements.
This manuscript: Incidence and epidemiologic features of dengue in Sabah, Malaysia, was published in the journal Plos Neglected Tropical Diseases, on 11 May, 2020. https://doi.org/10.1371/journal.pntd.0007504. Individual contributions of each author are noted_ within the manuscript.
Principal Supervisor Confirmation:
I have sighted email or other. correspondence from all Co-authors confirming their certifying authorship.
Dr Francesca D. Frentiu QUT Verified Signature Name Signature Date
Chapter 3: Results for Objective 1 47 3.1. Incidence and epidemiologic features of dengue in Sabah, Malaysia
Amanda K. Murphy1,2*, Giri Shan Rajahram3,4, Jenarun Jilip5, Marilyn Maluda5, Timothy William4,6, Wenbiao Hu7, Simon Reid8, Gregor J. Devine1^, Francesca D. Frentiu2^.
1 Mosquito Control Laboratory, QIMR Berghofer Medical Research Institute, Brisbane, Australia 2 School of Biomedical Sciences, and Institute for Health and Biomedical Innovation, Queensland University of Technology, Brisbane, Australia 3 Queen Elizabeth Hospital, Ministry of Health Malaysia, Kota Kinabalu, Malaysia 4 Infectious Disease Society of Kota Kinabalu‐Menzies School of Health Research Clinical Research Unit, Kota Kinabalu, Malaysia 5 Sabah Department of Health, Ministry of Health Malaysia, Kota Kinabalu, Malaysia 6 Gleneagles Kota Kinabalu Hospital Sabah, Kota Kinabalu, Malaysia 7 School of Public Health and Social Work, Queensland University of Technology, Brisbane, Australia 8 School of Public Health, University of Queensland, Brisbane, Australia
^ These authors contributed equally to this work
* Corresponding author:
Amanda Murphy: [email protected]
Keywords: dengue, rural, Sabah, Aedes albopictus, Borneo, South East Asia
Chapter 3: Results for Objective 1 48
Abstract In South East Asia, dengue epidemics have increased in size and geographical distribution in recent years. We examined the spatiotemporal distribution and epidemiological characteristics of reported dengue cases in the predominantly rural state of Sabah, in Malaysian Borneo–an area where sylvatic and urban circulation of pathogens are known to intersect. Using a public health data set of routinely notified dengue cases in Sabah between 2010 and 2016, we described demographic and entomological risk factors, both before and after a 2014 change in the clinical case definition for the disease. Annual dengue incidence rates were spatially variable over the 7‐year study period from 2010–2016 (state‐wide mean annual incidence of 21 cases/100,000 people; range 5‐42/100,000), but were highest in rural localities in the western districts of the state (Kuala Penyu, Nabawan, Tenom and Kota Marudu). Eastern districts exhibited lower overall dengue rates, although a high proportion of severe (haemorrhagic) dengue cases (44%) were focused in Sandakan and Tawau. Dengue incidence was highest for those aged between 10 and 29 years (24/100,000), and was slightly higher for males compared to females. Available vector surveillance data indicated that during large outbreaks in 2015 and 2016 the mosquito Aedes albopictus was more prevalent in both urban and rural households (House Index of 64%) than Ae. aegypti (15%). Demographic patterns remained unchanged both before and after the dengue case definition was changed; however, in the years following the change, reported case numbers increased substantially. Overall, these findings suggest that dengue outbreaks in Sabah are increasing in both urban and rural settings. Future studies to better understand the drivers of risk in specific age groups, genders and geographic locations, and to test the potential role of Ae. albopictus in transmission, may help target dengue prevention and control efforts.
Chapter 3: Results for Objective 1 49 Author summary In order to combat the rising regional incidence of dengue in South East Asia, incidence patterns must be better characterised within different ecological settings. We conducted the first retrospective analysis of dengue epidemiology in the predominantly rural state of Sabah, Malaysia, where both urban and sylvatic transmission cycles exist. Human notification data over a 7‐year period were reviewed and spatiotemporal and demographic risk factors identified. We found that:
1. While urban habitats do play a role in mediating the spread of dengue in Sabah, cases from both urban and rural localities contributed equally to dengue outbreaks; 2. Human demographic risk factors included being aged between 10 and 29 years, and being male;
3. Cases of severe dengue were more common in the eastern districts of the state. Almost half of severe dengue cases were reported from Sandakan and Tawau; and 4. The presence of Aedes albopictus in and around the majority of urban and rural case households, often in the apparent absence of Ae. aegypti, suggests that its role in transmission requires further investigation.
This study emphasises that increasing incidence of dengue in urban South East Asia is also mirrored in more rural areas, and suggests a need for control strategies that address both urban and rural dengue risk.
Chapter 3: Results for Objective 1 50
Introduction Dengue is the most rapidly spreading vector‐borne disease in the world, and the most prevalent arboviral disease of humans (1). Now endemic in more than 100 countries, the disease causes an enormous burden on communities and health care systems in tropical and sub‐tropical regions (2). The causative agent of dengue is dengue virus (DENV), transmitted between humans by Aedes mosquitoes across a range of domestic and sylvatic environments. Urban expansion, human migration, travel and trade have facilitated an increasing number of infections, primarily in Asia, Africa and North and South America (3, 4). The Americas and Asia have been identified as high risk zones due to the presence of dense populations of humans and vectors and climate suitability, with Asia in particular having a disproportionate burden (up to 70%) of global infections (1). The number, severity and geographic distribution of dengue epidemics has increased in South East Asia since the 1950s, when the first cases of the severe (haemorrhagic) form of dengue were identified during epidemics in Thailand and the Philippines (3, 5). Outbreaks continued throughout South East Asia during the 1960s, including Vietnam, the Philippines, Singapore, Malaysia and Thailand, with Malaysia’s first severe dengue outbreak recorded in Penang in 1962 (61 cases and 5 deaths reported) (6).
In Malaysia, regular outbreaks have occurred since the 1960s with the first major, nation‐wide outbreak of 1,487 cases and 54 deaths recorded in 1973 (7). Steep increases in case numbers began to occur from the late 1980s when incidence rates rose from 9 cases/100,000 in 1988 to 123 cases/100,000 in 1998, reaching a total of 27,381 reported cases in 1998 (8). These increases occurred alongside Malaysia’s rapid population and infrastructure growth (rising from 13.7 – 23.3 million people between 1980 and 2000), which facilitated the spread of dengue through the unintended creation of new vector breeding sites (9, 10). This trend continued into the 21st century, with a 7‐fold increase in case numbers between 2000 and 2010, when case numbers reached 46,171 (incidence rates of 30 and 162/100,000, respectively) (11). In 2014, the largest ever outbreak was recorded, with 108,698 cases (incidence rate of 361/100,000, with 215 deaths) (12, 13). This coincided with the introduction of a new national dengue case definition in that same year, requiring all notifications to be laboratory‐confirmed by a
Chapter 3: Results for Objective 1 51 diagnostic test (either NS1 and/or IgM/IgG serology) (14). While the majority of reported cases have been concentrated in the highly urbanized states of Selangor, Kuala Lumpur and Johor, located on the Malaysian peninsula (together comprising 68% of cases between 2012 and 2016), increases have also been recorded in more rural states (15, 16). Few eco‐epidemiological studies have explored the factors driving incidence rates in rural parts of the country; however, seroprevalence and vector surveillance studies suggest that infection risk has become comparable in rural and urban areas alike (16‐ 19).
As with many South East Asian countries, the drivers of dengue disease in Malaysia are multi‐faceted, and encompass characteristics of, and interactions between, virus, vectors, hosts and their environments. These include viral virulence and human biological factors; climate factors, including high temperature, relative humidity and increased rainfall; human movement and behaviour; and economic and infrastructure development (9, 20). These may alter human susceptibility to infection, promote mosquito breeding and/or increase interactions of viruses, vectors and hosts. The Malaysian Ministry of Health has reported inappropriate waste disposal to be a major dengue prevention challenge, with polystyrene food containers, plastic bottles and tyres contributing the highest percentage of artificial mosquito breeding sites (21). Mosquito breeding and human movement facilitate the ongoing circulation of all four human DENV serotypes across the country, although circulation patterns are distinct between states (22, 23). In addition, sylvatic DENV serotypes detected from human cases on the island of Borneo suggest a greater diversity of viruses in some habitats (24‐26) and the potential for sylvatic viruses to enter human transmission cycles (27).
The Malaysian states of Sabah and Sarawak, in Malaysian Borneo, report lower incidence rates than mainland Malaysia (together, 5% of all cases between 2012‐2016) and patterns of transmission in these states are not well characterised (15). Rapidly developing urban areas are located in close proximity to disturbed forest environments, facilitating interactions among DENV vectors and hosts, and increasing potential risk of spillover of sylvatic pathogens to human populations (24, 28). Sabah has high rates of forest loss, with monocultures of rubber and palm plantations estimated to cover 36‐ 56% of the land area [28, 29], and reports the highest incidence of the sylvatic malaria
Chapter 3: Results for Objective 1 52
parasite Plasmodium knowlesi [30]. Increased incidence rates of P. knowlesi malaria in Sabah have been linked to deforestation and land use changes, which lead to increased contact between its Anopheline vectors, primate hosts and humans [31,32]. Given the marked environmental change occurring in Sabah, and the increase in dengue cases noted in recent years (15, 26), it is essential from a public health perspective to understand current disease patterns and their associated risk factors. Our study describes recent spatial and temporal trends in dengue incidence, and some potential demographic and entomological risk factors from this understudied region of South East Asia.
Methods Ethics statement
This study was approved by the Medical Research and Ethics Committee (MREC), Ministry of Health Malaysia; and the Human Research Ethics Committee (HREC) of the QIMR Berghofer Medical Research Institute, Brisbane, Australia. All human case data analysed were anonymized.
Study site
The Malaysian state of Sabah lies at the most north‐eastern tip of the island of Borneo. It borders the Malaysian state of Sarawak and the Indonesian province of Kalimantan. The climate is tropical; there is high humidity and year‐round rainfall, which increases between November and March. Sabah has a geographical area of 73,904 km2 and is divided into 25 districts (32). Of Malaysia’s 13 states and 3 territories, Sabah’s population density is second lowest in the country (44 people/km2), after Sarawak (20 people/km2) (Fig. 3.1). Sabah also has a relatively low urban population proportion compared with other Malaysian states (54% of Sabah’s population live in urban areas, compared with 100% in Kuala Lumpur and Putrajaya) (15). Within Sabah, Kota Kinabalu district has the highest population density (1,397 people/km2, with almost 500,000 people), where the capital city of the same name is located.
Chapter 3: Results for Objective 1 53 Figure 3.1. Map of Malaysia and Sabah state.
Epidemiological data
State‐wide data from monthly notified cases of dengue between the years 2010 and 2016 were obtained from the Sabah State Department of Health (Jabatan Kesihatan Negeri, Sabah), Malaysian Ministry of Health. The data were collected as part of routine public health monitoring of dengue case reports from health facilities in Sabah. Prior to 2010, these data were not available in disaggregated and electronic format. Within the data set made available for this study, epidemiological variables included the age and gender of each case, the district and locality (urban or rural) of each case residence, disease severity and outcome (survival or death), and diagnostic tests performed (IgG, IgM and/or NS1). In our dataset, locality status was available for the 6‐year period from 2011‐2016. Locality of case residence represented the smallest residential geographical unit used by the Malaysian Department of Statistics, and was allocated based on home address. Designation of locality status (urban or rural) was according to the Malaysian Department of Statistics definitions of urban and rural, where urban localities are gazetted census areas with 10,000 people or more, and ≥60% of the working population (≥15 years) engaged in non‐agricultural activities (15). In our dataset, locality status was recorded for case residences from 2011 onwards. Population and demographic data for Sabah districts were obtained from the Malaysian Department of Statistics for the year 2010. Estimated population numbers for each age, gender and district were then calculated for each subsequent year using published annual population growth rates (10). These estimated resident population numbers were used to calculate mean annual incidence rates.
During the study period, clinical cases were identified using World Health Organization (WHO) guidelines using clinical symptoms and/or positive NS1 or serology (presence of IgM or IgG) (33). Diagnosis of severe dengue was based on identifying more serious symptoms including severe plasma leakage, severe haemorrhage and severe organ dysfunction (34). From 2014 onwards, Malaysian national notification guidelines were modified, in line with WHO advice, to require a positive laboratory diagnostic test (either NS1 and/or IgM/IgG serology) in addition to the presence of clinical symptoms,
Chapter 3: Results for Objective 1 54
and case notification within 24 hours of diagnosis (14, 35). Therefore, the majority of cases prior to 2014 were clinically diagnosed (with 30‐50% per year confirmed by laboratory tests in our dataset), while cases from 2014‐2016 were 100% laboratory confirmed.
Entomological data
Entomological surveillance data (number of larvae, mosquito species identified) were generated from active surveillance of potential aquatic habitats, primarily water‐ holding containers, indoors and outdoors around 719 case residences inspected during the 2015‐16 outbreaks. Surveys were carried out by the local public health authority who designated 255 (36%) of inspected residences as urban, and 437 (61%) as rural. No rural or urban designation recorded for a further 27 (4%) of houses. Where mosquito larvae were found in or around a case household, samples were taken to local public health laboratories for species identification. The presence or absence of one or more species per household was recorded, and the House Index (HI) was calculated as the proportion of houses infested with larvae and/or pupae (36).
Data analysis
We assessed seasonal characteristics of the temporal distribution of cases using a seasonal trend decomposition procedure in SPSS software. The procedure is based on the Census Method I, otherwise known as the ratio‐to‐moving‐average method where time series data are separated into a seasonal component, a combined trend and cycle component, and an "error" or irregular component (37). The seasonal component is then isolated from the overall and irregular trends through a multiplicative model. Seasonal decomposition analysis was applied to monthly dengue case numbers across the 7‐year period to examine the seasonal trends of case notifications across Sabah.
Annual, monthly and mean incidence of dengue were calculated using the number of notifications per month and Sabah population estimates based on the 2010 Malaysian census. Incidence rates were standardized for age and gender by adjusting population numbers for these variables by their relative district‐level population proportions. Ages of cases were grouped into four categories to broadly separate young children from older children and adults (0‐9, 10‐29, 30‐49 and ≥50). Differences between means were
Chapter 3: Results for Objective 1 55 determined by Kruskal‐Wallis or Mann‐Whitney statistical tests using SPSS Statistics software (SPSS, IBM New York USA; version 23). Statistical significance was set at p<0.05. Maps of Malaysia and Sabah dengue cases and incidence were created using ArcGIS (Esri Redlands USA; version 10.5.1).
We assessed overall and annual trends of rural versus urban cases (locality status) at the state‐wide level for a 6‐year period where this variable was available (2011‐2016). This included a total of 9,791 cases. Of these, 756 (7.7%) cases were missing a designated locality status (rural or urban), due to incomplete data entry. We excluded these from rural and urban incidence calculations. For the remaining 9,035 cases, we calculated the total proportions and incidence rates for urban and rural cases, using population projections calculated from state‐wide rural‐urban population data published in 2010 (10). At district level, we calculated annual and overall proportions of rural and urban cases per district. Where cases with unspecified localities were included in analyses (Tables 3.1, 3.2 and S3.3), the proportion of unspecified localities were indicated. Annual and overall relative risks (RR) of dengue for each individual district were calculated using:
Observed incidence rate RR Expected incidence rate
where the expected incidence rate for each district is based on the mean rate for the state multiplied by the population of each district. A RR value >1 indicates increased incidence of dengue in that location compared to the expected (mean) incidence, and a value <1 indicates lower than expected dengue incidence.
Results Temporal trends across the state
A total of 11,882 dengue cases were notified in Sabah during the 7‐year study period, with 25 deaths. Cases were notified year‐round, with outbreaks commonly occurring in the second half of the year between July and December, sometimes continuing into January and February (Fig. 3.2). Seasonal decomposition analysis showed that, on average, notifications peaked each January, with the highest risk period
Chapter 3: Results for Objective 1 56
being between November and March. The typical off‐peak months were between April‐ June (S3.1 Fig).
Figure 3.2. Temporal pattern of dengue in Sabah, 2010‐2016.
Outbreaks varied in magnitude between years, with the largest outbreaks in 2010 and from 2015‐2016 (Fig. 3.2). Between 2010‐2013, the mean annual incidence was 13 cases/100,000 and this increased to 32 cases/100,000 between 2014‐2016, coinciding with the change in national notification guidelines. The mean state‐wide annual incidence rate across the 7 years was 21 cases per 100,000 people (median 18/100,000). For the 6‐year period (2011‐2016) where locality data were available, state‐wide mean annual incidence of dengue in urban localities was 44/100,000 versus 47/100,000 for rural localities. Annual rates of dengue in urban and rural localities often contributed similarly to the overall burden, despite annual variations in incidence (Fig. 3.3). Their overall contributions to dengue incidence were also roughly equivalent both before and after the 2014 revised case definition.
Figure 3.3. State‐wide annual incidence of dengue in rural and urban localities, 2011‐2016.
Chapter 3: Results for Objective 1 57 Demographic trends
Analyses of demographic trends across Sabah indicated a slightly higher proportion of male dengue cases (60%) than females (40%); however, mean annual incidence rates were not significantly different between the two genders across the 7 years (29/100,000 for males and 20/100,000 for females, Mann‐Whitney U=19, p=0.535), nor within different age‐groups (U=321, p=0.241) (Fig. 3.4). This was relatively consistent across all Sabah districts; however, above‐average proportions of male cases were observed in Tongod and Kinabatangan (75% and 65% male cases, respectively).
Figure 3.4. Incidence of dengue in Sabah by age group and gender, 2010‐2016.
For both genders, nearly half the case burden (47%) was borne by those aged between 10 and 29 years (mean annual incidence of 24 cases/100,000), followed by 30‐ 49 years (26% of total cases, and 13 cases/100,000). The median age of all notifications was 25. After standardising for age and gender, incidence rates were found to be significantly different between age groups (Kruskal‐Wallis H=9.046, p=0.029), with pairwise comparisons identifying a significant difference between the 0‐9 and 10‐29 age groups only (Mann‐Whitney U=5.0, p=0.011). The lowest proportion of notifications occurred below 10 years of age (mean annual rate of 6 cases/100,000), followed by those ≥50 years of age (8 cases/100,000). The incidence trends for gender and age groups remained consistent both before and after the 2014 case definition change.
Spatial trends across districts
District‐level incidence rates were highly variable each year, with a mean annual rate of 50 cases/100,000 (range 19‐161/100,000) across the 7 years (Table 3.1, S3.2). High annual variability meant that there was no significant difference in mean incidence rates between the districts overall (H=25.978, p=0.354); however, the highest mean incidence rates were found in districts in the west of the state with relatively low human
Chapter 3: Results for Objective 1 58
population density, including Kuala Penyu, Nabawan, Tenom and Kota Marudu. The relative risks were highest in Kuala Penyu, Kota Marudu and Kudat districts (RR=3.5, 2.1 and 1.8, respectively) (Table 3.1). The 4 highest‐incidence districts reported a low proportion of cases residing in urban localities (0‐12%) (Table 3.1). Lower, less variable incidence rates were recorded from some of the central and eastern districts including Kinabatangan, Tongod, Kunak and Tawau (annual incidence range of 3‐62 cases/100,000 each year). These districts also had some of the lowest relative risks (RR=0.3, 0.4, 1, and 0.8, respectively) (Table 3.1), along with a wide range in proportions of urban cases (4‐ 73%).
The changing annual spatial trend is shown in Fig. 3.5, which indicates high annual variability across districts, and the highest mean incidence rates overall in the western districts of Sabah. A spatial shift occurred during the large 2015 outbreak, when incidence increased markedly in the more densely populated western districts of Kota Kinabalu, Penampan, and Putatan, as well as in Sandakan and Semporna in the east (Fig. 3.5). The overall urban case proportions in these districts ranged from 53‐85%. During 2016, cases from both urban and rural localities contributed to the outbreak, but the greatest overall burden was in rural localities of Tenom, Nabawan and Keningau districts.
Table 3.1. Summary of population and dengue burden across Sabah, 2010‐2016.
Figure 3.5. Annual spatial incidence of dengue in Sabah, 2010‐2016.
Severe dengue
Of all dengue cases reported over the 7 years, 1.1% were severe (haemorrhagic) dengue cases. Severe cases were reported across all years, with an average of 18 severe cases/year across the state. The greatest proportion of severe cases were concentrated in Sandakan (24%) and Tawau (20%) districts on the eastern side of the state, and this was consistent across all years. The highest severe dengue incidence occurred in Kunak,
Chapter 3: Results for Objective 1 59 Sandakan and Tongod, while the lowest was observed in the western districts, several of which recorded zero severe cases despite recording high overall dengue incidence (Figs. 3.5 and 3.6, Table 3.1). Severe dengue occurred in expected proportions relative to population size of both genders and age groups, with the highest burden (35%) reported in the 10‐19 years age group. There were 9 severe dengue deaths during the study period, 4 of which were in Tawau district.
Figure 3.6. Total severe dengue notifications by district, 2010‐2016.
Entomological factors
Entomological data (House Indices) collected during the 2015‐2016 outbreaks indicated that the partially sylvatic vector Aedes albopictus was the predominant species identified from larval collections in and around rural and urban case residences (Table 3.2). Of 719 dengue case residences that were inspected as part of active surveillance in 2015‐2016, 618 were found to contain mosquito larvae (HI=86%). Of those, Ae. albopictus larvae were identified from 394 residences (HI=64%), either alone (383 residences) or with Ae. aegypti (11 residences). Conversely, 94 residences were positive for Ae. aegypti (83 alone, 11 with Ae. albopictus; HI=15%), 33 residences were positive for Culex species (HI=5%), and 108 samples could not be identified (17%).
The specific districts where case residences were inspected are detailed in S3.3 Table. Seven hundred and nineteen inspections were conducted in 21/25 districts, with the greatest number performed in Tawau (158 inspections; 96 of which were in urban localities) and Nabawan (107 inspections; all in rural localities). The residences with the highest proportions of Ae. aegypti were in the east coast districts of Tawau (larvae found in 60/158 residences) and Lahad Datu (18/43 larvae‐positive residences). Ae. albopictus was the most prevalent species across both urban and rural residences of most of districts surveyed.
Chapter 3: Results for Objective 1 60
Table 3.2. Mosquito larvae species collected from case residences in Sabah during 2015‐ 2016.
Discussion Spatial and temporal trends
In recent decades, the scale of epidemics in South East Asia has increased, including in the Malaysian state of Sabah, where incidence rates have increased substantially since 2014. While spatial trends varied from year to year, the highest incidence across all years occurred in districts along the western coast of the state. The timing of large epidemic years in Sabah (2010, 2015 and 2016) was consistent with patterns observed at national and regional levels during the same period (wider Malaysia, Indonesia, Philippines) (11, 38). This suggests shared climatic influences on outbreak occurrence. Across the tropics, dengue transmission occurs all year round with temporal peaks and troughs whose causative relationship with temperature and rainfall remains poorly defined (39, 40).
Our observations are derived from district‐level incidence data and we were not able to evaluate finer scale incidence patterns. However, we were able to illustrate marked spatial and temporal heterogeneity across Sabah in all years and noted a distinct shift in the spatial dominance of urban versus rural localities during the large outbreak of 2015, which reversed again in 2016. Our findings suggest that high density, urbanised areas are not necessarily the primary locations of ongoing epidemics in Sabah. Similar observations of rural dominance or equivalence in dengue incidence have also been observed in other areas of Malaysia (16, 41) and elsewhere in the region, including in Cambodia, Thailand, Vietnam and Sri Lanka (42‐46). These latter countries have all reported epidemics that shift between rural and urban areas via human or mosquito movement. In endemic environments, transmission is mediated by water storage and waste disposal practices, mosquito vector ecology and sociocultural factors including human movement patterns (47‐51). The relative importance of these is still not well understood (52, 53). In Sabah, effectively addressing these risks will require dengue management strategies to be applied across a range of ecological settings.
Chapter 3: Results for Objective 1 61 Demographic factors
Our findings indicated that the age‐related dengue risk in Sabah was in line with regional trends indicating a transition from children to adults being disproportionately affected by dengue (54). Incidence was higher for males than for females across all districts of the state, and was significantly higher for both genders in the 10‐29 age group. This higher risk may suggest that a large proportion of people in this age‐group (and males in particular) were either engaged in outdoor activities or possibly being occupationally exposed to mosquitoes. Although exposure to dengue vectors can occur indoors, outdoor activities, especially those in close proximity to forests or forest edges, are thought to increase the risk of being bitten by the abundant exophilic vector, Ae. albopictus (55, 56).
The agriculture sector is the major employment sector in Sabah. Living or working in proximity to rubber plantations or forested areas has been linked to increased vector‐ borne disease risk, in Sabah and in other parts of South East Asia (57, 58). This is due to changing human interactions with particular land use types, where vector and host ecology have been altered (9, 55, 59). The particularly high proportion of males affected in the largely rural Tongod and Kinabatangan districts may reflect a greater proportion of men engaged in occupational or recreational outdoor activities. Further investigation of employees of outdoor occupations, such as agricultural workers, or of residential populations close to agricultural or forested areas of Sabah, could shed light on the specific demographic factors and land use characteristics associated with infection.
Severe dengue
Changing demographic or immunological factors may explain the observed pattern of severe dengue in our study. Incidence of severe dengue was localised to two main regions of the state: the eastern districts of Tawau and Sandakan. These districts include major urbanised cities as well as rural surrounding areas, and comprised relatively low dengue rates compared to the west of the state. The reasons for this spatial concentration of cases in these eastern districts is unknown. It is possible that the increase in severe cases in these areas followed a serotype switch from DENV 4 to DENV 1 reported to have occurred in Sandakan between 2013 and 2016 (26). Major dengue outbreaks commonly follow the switching of DENV serotypes and the
Chapter 3: Results for Objective 1 62
subsequent loss of herd immunity in the human population. Those with a history of dengue infection with a different serotype may be more vulnerable to severe dengue as a result of antibody‐dependent enhancement of infection (60‐62). Surveillance information regarding which virus serotypes and genotypes were circulating in Sabah was not available in this study, so we were unable to assess the potential contribution of virus circulation patterns to the trends we observed. However, routine public health monitoring of circulating serotypes, including the importation of new serotypes in particular, should be prioritised. Close monitoring of serological surveillance data alongside detailed epidemiological data could aid predictions of severe disease risk (63, 64).
Entomological factors
Ae. albopictus was by far the commonest potential dengue vector identified by the public health authorities. It was three times more common than Ae. aegypti in urban case households and seven times more common in rural households. In >50% of case households inspected, Ae. albopictus was the only potential vector identified. The eastern districts of Sabah state appeared to have a higher proportion of Ae. aegypti compared to the rest of the state, although overall dengue incidence was lower on the east coast. This finding was consistent with those of early entomological surveys of Sabah in the 1970’s, which reported higher numbers of Ae. aegypti on the east coast and lower abundance on the west coast (65, 66). In those studies, the greater presence of Ae. aegypti in the east was thought to be due to more frequent travel by boat between east coast settlements for fishing and trade. Although Ae. aegypti is generally considered responsible for most dengue transmission in South East Asia (67, 68), Ae. albopictus is more common than Ae. aegypti across Malaysia and Borneo (55, 56, 59, 69). Our results are consistent with these previous surveys, and indicate that this species was dominant during the two largest and most recent dengue outbreaks in our study period.
The presence of natural and artificial larval habitats for Ae. albopictus have previously been associated with epidemics in both urban and rural areas of Malaysia (70‐72). Urban dominance of Ae. albopictus has also been observed, at least seasonally, in parts of Thailand, southern China and other South East Asian countries (73‐75). Vector
Chapter 3: Results for Objective 1 63 incrimination studies for Malaysia are rare, but Ming et al. showed that Ae. albopictus was 3‐10 fold more abundant than Ae. aegypti around Kuala Lumpur and 3‐5 times less likely to harbour dengue virus (76). The potential role of Ae. albopictus in transmission in Sabah, and how that might be affected by human density and biting preference in rural areas, requires further investigation.
Limitations
The revised dengue case definition in Malaysia in 2014 will undoubtedly have influenced the incidence rates reported here, in terms of either under‐ or over‐reporting of cases. Less reliance on clinical symptoms for case notification from 2014 onwards might have been expected to reduce notifications but, in fact, a dramatic increase in cases was recorded. Perhaps the influx of diagnostic kits into health facilities across the country encouraged testing of all febrile cases, accompanied by changes in awareness and reporting by clinicians. Prior to 2014, there may have been a lack of resources for testing or notifying dengue, or other socioeconomic factors, that resulted in under‐ reporting (77). It is also possible that increases in diagnostic testing from 2014 were not uniform across all districts, and/or that additional reporting inconsistencies at the sub‐ district level, or between rural and urban health facilities, may have impacted our observations. However, widespread increases in cases recorded throughout the rest of the country during the same time frame suggest that much of the true dengue burden was not being captured prior to 2014. Longer‐term monitoring of notification trends will likely clarify the true burden. Similarly, finer‐scale investigation at the sub‐district level might enable more detailed analyses of dengue risk factors, as variation in population density and demographic factors within districts could be more accurately assessed. Additional immunological, biological and/or virological factors may also have impacted our findings, but were beyond the scope of this study to assess.
The use of house index (HI, percentage of houses positive for mosquito larvae) for entomological surveillance is not a good correlate of dengue infection risk, but it remains one of the most widely used entomological indices of operational programs. Larval surveys are, however, appropriate to survey for the presence/absence of potential vectors. In this study, the HIs for Ae. aegypti and Ae. albopictus in Sabah indicate that Ae. albopictus is extremely common around dengue case houses, often in the apparent
Chapter 3: Results for Objective 1 64
absence of Ae. aegypti. It remains possible that Ae. aegypti was simply more cryptic than Ae. albopictus and that, despite its apparent absence, it was the key vector. Nonetheless, the relative roles of Ae. albopictus and Ae. aegypti as vectors in Sabah are important to define through further vector incrimination studies.
Future directions
Our study describes spatial, temporal, demographic and vector‐related characteristics of dengue disease patterns in Sabah state. To further explain some of the trends observed, finer‐scale collection of demographic data, and additional field investigations are necessary. In particular, surveys of socioeconomic variables associated with dengue, including human movement in urban and rural landscapes, would aid identification of high‐risk groups. Further investigations and monitoring of the spatial and temporal movement of virus serotypes and vectors could also inform prevention strategies. Ultimately, linking both vector and serological surveillance to the dengue case notification system would enhance existing public health efforts. Considering the ongoing expansion of dengue endemicity and burden in the region, proactive strategies to increase understanding of the complex and evolving epidemiological factors underlying dengue risk across varied environments are critical.
Chapter 3: Results for Objective 1 65 Acknowledgements
The authors would like to acknowledge the contribution of the Sabah Department of Health, Ministry of Health, Malaysia for making dengue notification data available. We also thank the Director General of Health Malaysia for the permission to publish this paper. We are grateful for assistance and input provided by Prof. Nicholas Anstey, Dr Matthew Grigg, Dr Kimberley Fornace, Dr Christopher Wilkes, Dr Eloise Stephenson and Dr Andrea Rabellino.
Authors’ Contributions
AM, JJ and TW conceived the project. AM, JJ, and MM curated the data. AM carried out the analyses, with input from GSR, MM, WH, SR, FF and GD. AM drafted the paper. GSR, MM, TW, SR, GD and FF aided interpretation. All authors reviewed and revised the final manuscript and agreed to its submission.
Chapter 3: Results for Objective 1 66
References 1. Bhatt S, Gething PW, Brady OJ, Messina JP, Farlow AW, Moyes CL, et al. The global distribution and burden of dengue. Nature. 2013;496(7446):504‐7. doi: 10.1038/nature12060. PubMed PMID: 23563266; PubMed Central PMCID: PMCPMC3651993. 2. World Health Organization. Global Strategy for Dengue Prevention and Control, 2012‐2020. Geneva, 2012. 3. Kyle JL, Harris E. Global spread and persistence of dengue. Annu Rev Microbiol. 2008;62:71‐92. doi: 10.1146/annurev.micro.62.081307.163005. PubMed PMID: 18429680. 4. Messina JP, Brady OJ, Scott TW, Zou C, Pigott DM, Duda KA, et al. Global spread of dengue virus types: mapping the 70 year history. Trends Microbiol. 2014;22(3):138‐46. doi: 10.1016/j.tim.2013.12.011. PubMed PMID: 24468533; PubMed Central PMCID: PMCPMC3946041. 5. Vasilakis N, Weaver SC. The history and evolution of human dengue emergence. Adv Virus Res. 2008;72:1‐76. Epub 2008/12/17. doi: 10.1016/S0065‐ 3527(08)00401‐6. PubMed PMID: 19081488. 6. Halstead SB. Mosquito‐borne haemorrhagic fevers of South and South‐East Asia. Bull World Health Organ. 1966;35(1):3‐15. Epub 1966/01/01. PubMed PMID: 5297536; PubMed Central PMCID: PMC2476178. 7. Shekhar KC, Huat OL. Epidemiology of dengue/dengue hemorrhagic fever in Malaysia‐‐a retrospective epidemiological study. 1973‐1987. Part II: Dengue fever (DF). Asia Pac J Public Health. 1992;6(3):126‐33. Epub 1992/01/01. PubMed PMID: 1342799. 8. Hussin N, Jaafar J, Naing NN, Mat HA, Muhamad AH, Mamat MN. A review of dengue fever incidence in Kota Bharu, Kelantan, Malaysia during the years 1998‐ 2003. Southeast Asian J Trop Med Public Health. 2005;36(5):1179‐86. Epub 2006/01/28. PubMed PMID: 16438143. 9. Kwa B. Environmental change, development and vectorborne disease: Malaysia's experience with filariasis, scrub typhus and dengue. Environ Dev Sustain. 2008;10:209‐17. Epub 9 July 2006. PubMed PMID: DOI 10.1007/s10668‐006‐9060‐ 5. 10. Malaysian Department of Statistics. Population Distribution and Basic Demographic Characteristics, Population and Housing Census of Malaysia 2010. Kuala Lumpur: 2011. 11. Mia MS, Begum RA, Er AC, Abidin RD, Pereira JJ. Trends of dengue infections in Malaysia, 2000‐2010. Asian Pac J Trop Med. 2013;6(6):462‐6. Epub 2013/05/29. doi: 10.1016/S1995‐7645(13)60075‐9. PubMed PMID: 23711707. 12. Malaysian Remote Sensing Agency (ARSM). iDengue For Communities information system: Ministry of Health, Malaysia; 2013 [Accessed September 10, 2019]. Available from: http://idengue.arsm.gov.my/index.php. 13. World Health Organization, Western Pacific Region (WPRO). Health Information and Intelligence Platform: WPRO; 2014 [Accessed September 9, 2019]. Available from: https://hiip.wpro.who.int/portal/DataAnalytics/GenerateTables.aspx. 14. Malaysian Ministry of Health. Guidelines for Health Management and Disease Control. Putrajaya, Malaysia, 2013.
Chapter 3: Results for Objective 1 67 15. Malaysian Department of Statistics. Labour Force and Social Statistics. 2018. Available from: https://www.dosm.gov.my/v1/index.php?r=column/cthemeByCat&cat=152&bul_i d=aktTTjhhRHd1aHBCZGF1N01aaTl3dz09&menu_id=U3VPMldoYUxzVzFaYmNkWX ZteGduZz09. 16. Chew CH, Woon YL, Amin F, Adnan TH, Abdul Wahab AH, Ahmad ZE, et al. Rural‐ urban comparisons of dengue seroprevalence in Malaysia. BMC Public Health. 2016;16(1):824. doi: 10.1186/s12889‐016‐3496‐9. PubMed PMID: 27538986; PubMed Central PMCID: PMCPMC4991088. 17. Cheah WL, Chang MS, Wang YC. Spatial, environmental and entomological risk factors analysis on a rural dengue outbreak in Lundu District in Sarawak, Malaysia. Trop Biomed. 2006;23(1):85‐96. Epub 2006/10/17. PubMed PMID: 17041556. 18. Muhammad Azami NA, Salleh SA, Neoh HM, Syed Zakaria SZ, Jamal R. Dengue epidemic in Malaysia: Not a predominantly urban disease anymore. BMC Res Notes. 2011;4:216. Epub 2011/07/01. doi: 10.1186/1756‐0500‐4‐216. PubMed PMID: 21714858; PubMed Central PMCID: PMCPMC3154160. 19. Lim KW, Sit NW, Norzahira R, Sing KW, Wong HM, Chew HS, et al. Dengue vector surveillance in insular settlements of Pulau Ketam, Selangor, Malaysia. Trop Biomed. 2010;27(2):185‐92. PubMed PMID: 20962714. 20. Ahmad R, Suzilah I, Wan Najdah WMA, Topek O, Mustafakamal I, Lee HL. Factors determining dengue outbreak in Malaysia. PLoS One. 2018;13(2):e0193326. doi: 10.1371/journal.pone.0193326. PubMed PMID: 29474401. 21. Mudin RN. Dengue Incidence and the Prevention and Control Program in Malaysia. The International Medical Journal Malaysia. 2015;14(1):5‐9. 22. Mohd‐Zaki AH, Brett J, Ismail E, L'Azou M. Epidemiology of dengue disease in Malaysia (2000‐2012): a systematic literature review. PLoS Negl Trop Dis. 2014;8(11):e3159. Epub 2014/11/07. doi: 10.1371/journal.pntd.0003159. PubMed PMID: 25375211; PubMed Central PMCID: PMCPMC4222702. 23. Ng LC, Chem YK, Koo C, Mudin RNB, Amin FM, Lee KS, et al. 2013 dengue outbreaks in Singapore and Malaysia caused by different viral strains. Am J Trop Med Hyg. 2015;92(6):1150‐5. doi: 10.4269/ajtmh.14‐0588. PubMed PMID: 25846296; PubMed Central PMCID: PMCPMC4458818. 24. Vasilakis N, Cardosa J, Hanley KA, Holmes EC, Weaver SC. Fever from the forest: prospects for the continued emergence of sylvatic dengue virus and its impact on public health. Nat Rev Microbiol. 2011;9(7):532‐41. doi: 10.1038/nrmicro2595. PubMed PMID: 21666708; PubMed Central PMCID: PMCPMC3321645. 25. Liu W, Pickering P, Duchene S, Holmes EC, Aaskov JG. Highly Divergent Dengue Virus Type 2 in Traveler Returning from Borneo to Australia. Emerg Infect Dis. 2016;22(12):2146‐8. doi: 10.3201/eid2212.160813. PubMed PMID: 27869598; PubMed Central PMCID: PMCPMC5189156. 26. Gintarong TJ, Emran A, Sherin A, Thein TT, Aung TS. Circulation of all dengue virus serotypes during dengue outbreak in Sandakan, Sabah, Malaysia (2016). J Vector Borne Dis. 2018;55(2):168‐71. doi: 10.4103/0972‐9062.242566. PubMed PMID: 30280717. 27. Chen R, Vasilakis N. Dengue‐‐quo tu et quo vadis? Viruses. 2011;3(9):1562‐608. Epub 2011/10/14. doi: 10.3390/v3091562. PubMed PMID: 21994796; PubMed Central PMCID: PMCPMC3187692.
Chapter 3: Results for Objective 1 68
28. Gaveau DL, Sloan S, Molidena E, Yaen H, Sheil D, Abram NK, et al. Four decades of forest persistence, clearance and logging on Borneo. PLoS One. 2014;9(7):e101654. doi: 10.1371/journal.pone.0101654. PubMed PMID: 25029192; PubMed Central PMCID: PMCPMC4100734. 29. Malaysian Department of Statistics. Compendium of Environment Statistics. 2017. Available from: https://newss.statistics.gov.my/newss‐ portalx/ep/epFreeDownloadContentSearch.seam?cid=181040. 30. Cooper DJ, Rajahram GS, William T, Jelip J, Mohammad R, Benedict J, et al. Plasmodium knowlesi malaria in Sabah, Malaysia, 2015‐2017: ongoing increase in incidence despite near‐elimination of the human‐only Plasmodium species. Clin Infect Dis. 2019. doi: 10.1093/cid/ciz237. PubMed PMID: 30889244. 31. Fornace KM, Abidin TR, Alexander N, Brock P, Grigg MJ, Murphy A, et al. Association between Landscape Factors and Spatial Patterns of Plasmodium knowlesi Infections in Sabah, Malaysia. Emerging infectious diseases. 2016;22(2):201‐8. doi: 10.3201/eid2202.150656. PubMed PMID: 26812373; PubMed Central PMCID: PMCPMC4734530. 32. Brock PM, Fornace KM, Grigg MJ, Anstey NM, William T, Cox J, et al. Predictive analysis across spatial scales links zoonotic malaria to deforestation. Proc Biol Sci. 2019;286(1894):20182351. doi: 10.1098/rspb.2018.2351. PubMed PMID: 30963872; PubMed Central PMCID: PMCPMC6367187. 33. Organization WH. Dengue guidelines for diagnosis, treatment, prevention and control. In: (TDR) SPfRaTiTD, editor. France: WHO, Geneva, Switzerland; 2009. 34. WHO. Dengue Guidelines for Diagnosis, Treatment, Prevention and Control. Geneva: 2009. 35. World Health Organization (WHO). Handbook for clinical management of dengue. Geneva, Switzerland: WHO; 2012. 36. World Health Organization (WHO). Dengue control, vector surveillance: WHO; 2019 [Accessed April 4, 2019]. Available from: https://www.who.int/denguecontrol/monitoring/vector_surveillance/en/. 37. Macaulay FR. The Smoothing of Time Series. New York: The National Bureau of Economic Research; 1931. 38. Arima Y, Edelstein ZR, Han HK, Matsui T. Epidemiologic update on the dengue situation in the Western Pacific Region, 2011. Western Pac Surveill Response J. 2013;4(2):47‐54. Epub 2013/09/10. doi: 10.5365/WPSAR.2012.3.4.019. PubMed PMID: 24015372; PubMed Central PMCID: PMCPMC3762964. 39. Servadio JL, Rosenthal SR, Carlson L, Bauer C. Climate patterns and mosquito‐ borne disease outbreaks in South and Southeast Asia. J Infect Public Health. 2017. Epub 2017/12/25. doi: 10.1016/j.jiph.2017.12.006. PubMed PMID: 29274851. 40. de Castro DB, Sampaio VS, de Albuquerque BC, Pinto RC, Sadahiro M, Dos Passos RA, et al. Dengue epidemic typology and risk factors for extensive epidemic in Amazonas state, Brazil, 2010‐2011. BMC Public Health. 2018;18(1):356. doi: 10.1186/s12889‐018‐5251‐x. PubMed PMID: 29544456; PubMed Central PMCID: PMCPMC5855995. 41. Dhanoa A, Hassan SS, Jahan NK, Reidpath DD, Fatt QK, Ahmad MP, et al. Seroprevalence of dengue among healthy adults in a rural community in Southern Malaysia: a pilot study. Infect Dis Poverty. 2018;7(1):1. doi: 10.1186/s40249‐017‐ 0384‐1. PubMed PMID: 29335021; PubMed Central PMCID: PMCPMC5769361.
Chapter 3: Results for Objective 1 69 42. Rabaa MA, Ty Hang VT, Wills B, Farrar J, Simmons CP, Holmes EC. Phylogeography of recently emerged DENV‐2 in southern Viet Nam. PLoS Negl Trop Dis. 2010;4(7):e766. Epub 2010/07/30. doi: 10.1371/journal.pntd.0000766. PubMed PMID: 20668540; PubMed Central PMCID: PMC2910671. 43. Teurlai M, Huy R, Cazelles B, Duboz R, Baehr C, Vong S. Can human movements explain heterogeneous propagation of dengue fever in Cambodia? PLoS Negl Trop Dis. 2012;6(12):e1957. Epub 2012/12/14. doi: 10.1371/journal.pntd.0001957. PubMed PMID: 23236536; PubMed Central PMCID: PMCPMC3516584. 44. Thongsripong P, Green A, Kittayapong P, Kapan D, Wilcox B, Bennett S. Mosquito vector diversity across habitats in central Thailand endemic for dengue and other arthropod‐borne diseases. PLoS Negl Trop Dis. 2013;7(10):e2507. Epub 2013/11/10. doi: 10.1371/journal.pntd.0002507. PubMed PMID: 24205420; PubMed Central PMCID: PMCPMC3814347. 45. Bodinayake CK, Tillekeratne LG, Nagahawatte A, Devasiri V, Kodikara Arachichi W, Strouse JJ, et al. Emergence of Epidemic Dengue‐1 Virus in the Southern Province of Sri Lanka. PLoS Negl Trop Dis. 2016;10(10):e0004995. Epub 2016/10/07. doi: 10.1371/journal.pntd.0004995. PubMed PMID: 27711206; PubMed Central PMCID: PMCPMC5053469. 46. Vong S, Khieu V, Glass O, Ly S, Duong V, Huy R, et al. Dengue incidence in urban and rural Cambodia: results from population‐based active fever surveillance, 2006‐ 2008. PLoS Negl Trop Dis. 2010;4(11):e903. Epub 2010/12/15. doi: 10.1371/journal.pntd.0000903. PubMed PMID: 21152061; PubMed Central PMCID: PMC2994922. 47. Schmidt WP, Suzuki M, Thiem VD, White RG, Tsuzuki A, Yoshida LM, et al. Population density, water supply, and the risk of dengue fever in Vietnam: cohort study and spatial analysis. PLoS Med. 2011;8(8):e1001082. Epub 2011/09/16. doi: 10.1371/journal.pmed.1001082. PubMed PMID: 21918642; PubMed Central PMCID: PMCPMC3168879. 48. Arunachalam N, Tana S, Espino F, Kittayapong P, Abeyewickreme W, Wai KT, et al. Eco‐bio‐social determinants of dengue vector breeding: a multicountry study in urban and periurban Asia. Bull World Health Organ. 2010;88(3):173‐84. Epub 2010/04/30. doi: 10.2471/BLT.09.067892. PubMed PMID: 20428384; PubMed Central PMCID: PMC2828788. 49. Banu S, Hu W, Hurst C, Tong S. Dengue transmission in the Asia‐Pacific region: impact of climate change and socio‐environmental factors. Trop Med Int Health. 2011;16(5):598‐607. Epub 2011/02/16. doi: 10.1111/j.1365‐3156.2011.02734.x. PubMed PMID: 21320241. 50. Cheong YL, Leitao PJ, Lakes T. Assessment of land use factors associated with dengue cases in Malaysia using Boosted Regression Trees. Spat Spatiotemporal Epidemiol. 2014;10:75‐84. Epub 2014/08/13. doi: 10.1016/j.sste.2014.05.002. PubMed PMID: 25113593. 51. Rohani A, Suzilah I, Malinda M, Anuar I, Mohd Mazlan I, Salmah Maszaitun M, et al. Aedes larval population dynamics and risk for dengue epidemics in Malaysia. Trop Biomed. 2011;28(2):237‐48. Epub 2011/11/02. PubMed PMID: 22041742. 52. Anders KL, Nga le H, Thuy NT, Ngoc TV, Tam CT, Tai LT, et al. Households as foci for dengue transmission in highly urban Vietnam. PLoS Negl Trop Dis.
Chapter 3: Results for Objective 1 70
2015;9(2):e0003528. Epub 2015/02/14. doi: 10.1371/journal.pntd.0003528. PubMed PMID: 25680106; PubMed Central PMCID: PMCPMC4332484. 53. Ali M, Wagatsuma Y, Emch M, Breiman RF. Use of a geographic information system for defining spatial risk for dengue transmission in Bangladesh: role for Aedes albopictus in an urban outbreak. Am J Trop Med Hyg. 2003;69(6):634‐40. Epub 2004/01/27. PubMed PMID: 14740881. 54. Cummings DA, Iamsirithaworn S, Lessler JT, McDermott A, Prasanthong R, Nisalak A, et al. The impact of the demographic transition on dengue in Thailand: insights from a statistical analysis and mathematical modeling. PLoS Med. 2009;6(9):e1000139. Epub 2009/09/02. doi: 10.1371/journal.pmed.1000139. PubMed PMID: 19721696; PubMed Central PMCID: PMC2726436. 55. Brown R, Hing CT, Fornace K, Ferguson HM. Evaluation of resting traps to examine the behaviour and ecology of mosquito vectors in an area of rapidly changing land use in Sabah, Malaysian Borneo. Parasit Vectors. 2018;11(1):346. doi: 10.1186/s13071‐018‐2926‐1. PubMed PMID: 29898780; PubMed Central PMCID: PMCPMC6000972. 56. Rozilawati H, Tanaselvi K, Nazni WA, Mohd Masri S, Zairi J, Adanan CR, et al. Surveillance of Aedes albopictus Skuse breeding preference in selected dengue outbreak localities, peninsular Malaysia. Trop Biomed. 2015;32(1):49‐64. PubMed PMID: 25801254. 57. Tangena JA, Thammavong P, Lindsay SW, Brey PT. Risk of exposure to potential vector mosquitoes for rural workers in Northern Lao PDR. PLoS Negl Trop Dis. 2017;11(7):e0005802. Epub 2017/07/26. doi: 10.1371/journal.pntd.0005802. PubMed PMID: 28742854; PubMed Central PMCID: PMCPMC5544251. 58. Jeffree SM, Ahmed K, Safian N, Hassan R, Mihat O, Lukman KA, et al. Falciparum Malaria Outbreak in Sabah Linked to an Immigrant Rubber Tapper. Am J Trop Med Hyg. 2018;98(1):45‐50. Epub 2017/11/17. doi: 10.4269/ajtmh.17‐0081. PubMed PMID: 29141714; PubMed Central PMCID: PMCPMC5928689. 59. Young KI, Mundis S, Widen SG, Wood TG, Tesh RB, Cardosa J, et al. Abundance and distribution of sylvatic dengue virus vectors in three different land cover types in Sarawak, Malaysian Borneo. Parasit Vectors. 2017;10(1):406. doi: 10.1186/s13071‐ 017‐2341‐z. PubMed PMID: 28859676; PubMed Central PMCID: PMCPMC5580228. 60. Gubler DJ. Dengue and dengue hemorrhagic fever. Clin Microbiol Rev. 1998;11(3):480‐96. Epub 1998/07/17. PubMed PMID: 9665979; PubMed Central PMCID: PMC88892. 61. Lee KS, Lai YL, Lo S, Barkham T, Aw P, Ooi PL, et al. Dengue virus surveillance for early warning, Singapore. Emerg Infect Dis. 2010;16(5):847‐9. doi: 10.3201/eid1605.091006. PubMed PMID: 20409381; PubMed Central PMCID: PMCPMC2953985. 62. Suppiah J, Ching SM, Amin‐Nordin S, Mat‐Nor LA, Ahmad‐Najimudin NA, Low GK, et al. Clinical manifestations of dengue in relation to dengue serotype and genotype in Malaysia: A retrospective observational study. PLoS Negl Trop Dis. 2018;12(9):e0006817. doi: 10.1371/journal.pntd.0006817. PubMed PMID: 30226880; PubMed Central PMCID: PMCPMC6161924. 63. Fried JR, Gibbons RV, Kalayanarooj S, Thomas SJ, Srikiatkhachorn A, Yoon IK, et al. Serotype‐specific differences in the risk of dengue hemorrhagic fever: an analysis
Chapter 3: Results for Objective 1 71 of data collected in Bangkok, Thailand from 1994 to 2006. PLoS Negl Trop Dis. 2010;4(3):e617. Epub 2010/03/09. doi: 10.1371/journal.pntd.0000617. PubMed PMID: 20209155; PubMed Central PMCID: PMC2830471. 64. Tan KK, Zulkifle NI, Abd‐Jamil J, Sulaiman S, Yaacob CN, Azizan NS, et al. Disruption of predicted dengue virus type 3 major outbreak cycle coincided with switching of the dominant circulating virus genotype. Infect Genet Evol. 2017;54:271‐5. Epub 2017/07/13. doi: 10.1016/j.meegid.2017.07.008. PubMed PMID: 28698156. 65. Macdonald WW, Rajapaksa N. A survey of the distribution and relative prevalence of Aedes aegypti in Sabah, Brunei, and Sarawak. Bull World Health Organ. 1972;46(2):203‐9. PubMed PMID: 4537482; PubMed Central PMCID: PMCPMC2480700. 66. King JHL. A re‐survey of Aedes aegypti and Aedes albopictus in Sabah, Malaysia. Health Studies in Indonesia. 1978;11(1):45‐51. 67. Whitehorn J, Kien DT, Nguyen NM, Nguyen HL, Kyrylos PP, Carrington LB, et al. Comparative Susceptibility of Aedes albopictus and Aedes aegypti to Dengue Virus Infection After Feeding on Blood of Viremic Humans: Implications for Public Health. J Infect Dis. 2015;212(8):1182‐90. doi: 10.1093/infdis/jiv173. PubMed PMID: 25784733; PubMed Central PMCID: PMCPMC4577038. 68. Brady OJ, Golding N, Pigott DM, Kraemer MU, Messina JP, Reiner RC, Jr., et al. Global temperature constraints on Aedes aegypti and Ae. albopictus persistence and competence for dengue virus transmission. Parasit Vectors. 2014;7:338. doi: 10.1186/1756‐3305‐7‐338. PubMed PMID: 25052008; PubMed Central PMCID: PMCPMC4148136. 69. Black WCt, Hawley WA, Rai KS, Craig GB, Jr. Breeding structure of a colonizing species: Aedes albopictus (Skuse) in peninsular Malaysia and Borneo. Heredity (Edinb). 1988;61 ( Pt 3):439‐46. PubMed PMID: 3230033. 70. Chen CD, Nazni WA, Lee HL, Seleena B, Mohd Masri S, Chiang YF, et al. Mixed breeding of Aedes aegypti (L.) and Aedes albopictus Skuse in four dengue endemic areas in Kuala Lumpur and Selangor, Malaysia. Trop Biomed. 2006;23(2):224‐7. Epub 2007/02/27. PubMed PMID: 17322826. 71. Dieng H, Saifur RG, Hassan AA, Salmah MR, Boots M, Satho T, et al. Indoor‐ breeding of Aedes albopictus in northern peninsular Malaysia and its potential epidemiological implications. PLoS One. 2010;5(7):e11790. doi: 10.1371/journal.pone.0011790. PubMed PMID: 20668543; PubMed Central PMCID: PMCPMC2910701. 72. Roslan MA, Shafie A, Ngui R, Lim YA, Sulaiman WY. Vertical infestation of the dengue vectors Aedes aegypti and Aedes albopictus in apartments in Kuala Lumpur, Malaysia. J Am Mosq Control Assoc. 2013;29(4):328‐36. Epub 2014/02/21. PubMed PMID: 24551965. 73. Li Y, Kamara F, Zhou G, Puthiyakunnon S, Li C, Liu Y, et al. Urbanization increases Aedes albopictus larval habitats and accelerates mosquito development and survivorship. PLoS Negl Trop Dis. 2014;8(11):e3301. Epub 2014/11/14. doi: 10.1371/journal.pntd.0003301. PubMed PMID: 25393814; PubMed Central PMCID: PMCPMC4230920. 74. Peng HJ, Lai HB, Zhang QL, Xu BY, Zhang H, Liu WH, et al. A local outbreak of dengue caused by an imported case in Dongguan China. BMC Public Health.
Chapter 3: Results for Objective 1 72
2012;12:83. doi: 10.1186/1471‐2458‐12‐83. PubMed PMID: 22276682; PubMed Central PMCID: PMCPMC3311058. 75. Xu G, Dong H, Shi N, Liu S, Zhou A, Cheng Z, et al. An outbreak of dengue virus serotype 1 infection in Cixi, Ningbo, People's Republic of China, 2004, associated with a traveler from Thailand and high density of Aedes albopictus. Am J Trop Med Hyg. 2007;76(6):1182‐8. PubMed PMID: 17556633. 76. Ming LM, Adnan TH, Ooi CH, Ibrahim M, Husin D, Abdullah NA, et al. Adult Aedes Mosquito and Dengue Virus Surveillance in Residential and Public Areas of Selangor, Malaysia. Southeast Asian J Trop Med Public Health. 2018;49(4):553‐65. 77. Beatty ME, Stone A, Fitzsimons DW, Hanna JN, Lam SK, Vong S, et al. Best practices in dengue surveillance: a report from the Asia‐Pacific and Americas Dengue Prevention Boards. PLoS Negl Trop Dis. 2010;4(11):e890. Epub 2010/11/26. doi: 10.1371/journal.pntd.0000890. PubMed PMID: 21103381; PubMed Central PMCID: PMCPMC2982842.
Chapter 3: Results for Objective 1 73 Tables and Figures
Chapter 3: Results for Objective 1 74
Figure 3.1. Map of Malaysia and Sabah state.
Peninsula Malaysia and Malaysian Borneo are shown, along with the 13 Malaysian states and 3 territories: Kuala Lumpur (KL), Putrajaya (PJ) and Labuan. States are coloured according to their population density, expressed as number of people per square km. The island of Borneo includes the Malaysian states Sabah and Sarawak, and is also shared by the country of Brunei and the Indonesian province of Kalimantan. Inset: Sabah state, showing its 25 districts: Beaufort (BF), Beluran (BL), Keningau (KG), Kinabatangan (KT), Kota Belud (KB), Kota Kinabalu (KK), Kota Marudu (KM), Kuala Penyu (KP), Kudat (KD), Kunak (KN), Lahad Datu (LD), Nabawan (NB), Papar (PP), Penampang (PN), Pitas (PT), Putatan (PU), Ranau (RN), Sandakan (SD), Semporna (SM), Sipitang (SP), Tambunan (TB), Tawau (TW), Tenom (TN), Tongod (TG), Tuaran (TR). The three largest cities in the state are indicated by black points: the capital city Kota Kinabalu, Sandakan and Tawau.
Chapter 3: Results for Objective 1 75
Figure 3.2. Temporal pattern of dengue in Sabah, 2010‐2016.
The monthly number of reported dengue cases per year are shown (primary vertical axis), and the corresponding monthly incidence rate (secondary vertical axis). Grey bars indicate the years 2010–2013 where case diagnoses were predominantly clinically‐based (with or without laboratory confirmation), and blue bars indicate notifications following the 2014 change in case definition, where all cases were laboratory confirmed.
Chapter 3: Results for Objective 1 76
Figure 3.3. State‐wide annual incidence of dengue in rural and urban localities, 2011‐ 2016.
Annual incidence rates (per 100,000 population) across the state are shown for cases residing in either urban or rural localities, over a 6‐year period.
Chapter 3: Results for Objective 1 77
Figure 3.4. Incidence of dengue in Sabah by age group and gender, 2010‐2016.
Age‐ and gender‐adjusted incidence rates (per 100,000 population) across Sabah during the 7‐year period are shown for males, females and for both genders.
Chapter 3: Results for Objective 1 78
Figure 3.5. Annual spatial incidence of dengue in Sabah, 2010‐2016.
Incidence rates across districts are shown for each year, as well as the overall mean annual incidence during the 7‐year period. The 3 major cities of Sabah (Kota Kinabalu, Sandakan and Tawau) are indicated by black points.
Chapter 3: Results for Objective 1 79
Table 3.1. Summary of population and dengue burden across Sabah, 2010‐2016. Proportion of Severe dengue Human Population Mean annual cases from mean annual Overall relative Number of District population density incidence urban incidence risk dengue deaths (2010) (people/km2) (per 100,000) localities* (per 100,000) Beaufort 66,406 38 0.06 49 0.2 0.9 0 Beluran# 106,632 14 0.14 34 0.9 0.8 1 Keningau 177,735 50 0.11 57 0.2 0.9 0 Kinabatangan 150,327 23 0.04 19 0.0 0.3 0 Kota Belud 93,180 67 0.02 52 0.7 1.4 0 Kota Kinabalu 462,963 1,315 0.85 57 0.2 1 5 Kota Marudu 68,289 36 0.02 79 0.4 2.1 1 Kuala Penyu 19,426 43 0.06 161 0.0 3.5 0 Kudat 85,404 66 0.50 52 0.5 1.8 1 Kunak 62,851 55 0.62 33 1.3 1 0 Lahad Datu 206,861 28 0.42 45 0.8 1.3 2 Nabawan^ 32,309 5 0.00 88 0.0 1.6 0 Papar 128,434 103 0.20 31 0.0 0.6 0 Penampang 125,913 270 0.51 74 0.6 1.3 1 Pitas 38,764 27 0.29 36 0.0 0.8 0 Putatan** 55,864 1,397 0.32 73 0.0 1 2 Ranau 95,800 26 0.09 37 0.1 0.7 1 Sandakan 409,056 180 0.80 57 1.0 1.1 4 Semporna 137,868 120 0.53 46 1.1 1 2 Sipitang 35,764 13 0.14 56 0.0 1.2 0 Tambunan 36,297 27 0.00 37 0.0 0.8 0 Tawau 412,375 67 0.73 33 0.8 0.8 5 Tenom 56,597 13 0.12 84 0.7 1.5 0 Tongod** 36,192 4 0.05 22 1.0 0.4 0 Tuaran 105,435 90 0.26 56 0.4 1.3 0 Total 3,206,742 43 0.48 50 0.5 25 * Rural urban locality data was included from 2011‐2016; overall proportions were 0.48 urban, 0.44 rural and 0.08 unspecified. # Beluran was formerly known as Labuk Sugut. ^ Nabawan was formerly known as Pensiangan. ** Putatan and Tongod districts only commenced notifications in 2012.
Chapter 3: Results for Objective 1 80
Figure 3.6. Total severe dengue notifications by district, 2010‐2016.
Total number of severe dengue cases reported for each district during the 7‐year period. The 3 major cities of Sabah (Kota Kinabalu, Sandakan and Tawau) are indicated.
Chapter 3: Results for Objective 1 81
Table 3.2. Mosquito larvae species collected from case residences in Sabah during 2015‐2016.
Total No. of Number of larvae‐positive residences with specific species present (HI) number of No. of case larvae‐ House Locality dengue residences positive case Ae. Ae. Ae. aegypti & Index (HI) Culex spp. Undetermined cases in inspected residences aegypti albopictus Ae. albopictus* 2015‐2016 (all species) Urban residences 3,157 255 206 0.81 47 (0.23) 142 (0.69) 7 (0.03) 6 (0.03) 4 (0.02) Rural residences 2,824 437 388 0.89 33 (0.09) 225 (0.58) 4 (0.01) 26 (0.07) 100 (0.26) Locality unspecified 565 27 24 0.89 3 (0.13) 16 (0.67) 0 (0.0) 1 (0.04) 4 (0.17) Total 6,546 719 618 0.86 83 (0.13) 383 (0.62) 11 (0.02) 33 (0.05) 108 (0.17) HI = proportion of residences positive for mosquito larvae, calculated as number of residences with larvae/number of residences inspected. * Both species found breeding together in one household. Species were undetermined if the larvae failed to survive to adults to be identified, or if identification was pending/incomplete.
Chapter 3: Results for Objective 1 82
Supporting Information
S3.1 Figure. Seasonal decomposition of incidence rates in Sabah, 2010‐2016.
S3.2 Figure. Variation in dengue incidence across Sabah districts, 2010‐2016.
S3.3 Table. Entomological surveillance of case residences by district, 2015‐ 2016.
Chapter 3: Results for Objective 1 83
S3.1 Figure. Seasonal decomposition of incidence rates in Sabah, 2010‐2016.
The seasonal trend of dengue is shown in panel A, with the largest seasonal peak occurring on average between Nov and May each year (indicated by vertical black lines). The additional components separated from the seasonal trend during the decomposition procedure are also indicated in panels B‐D (cyclical component (B), irregular component (C) and overall smoothed trend (D)).
Chapter 3: Results for Objective 1 84
S3.2 Figure. Variation in dengue incidence across Sabah districts, 2010‐2016.
Dengue mean monthly incidence rates are plotted a) for the years 2010‐2013, and b) during 2014‐2016. The monthly mean (line), range (upper and lower whiskers), and outlying values are indicated for each district.
Chapter 3: Results for Objective 1 85
S3.3 Table. Entomological surveillance of case residences by district, 2015‐2016. Dengue case residence inspections Number of larvae‐positive residences with specific species present Total larvae‐ Ae. Urban Rural Unspecified Total positive Ae. Ae. aegypti & District Culex spp. Undetermined residences residences residences inspected residences aegypti albopictus Ae. (HI) albopictus Beaufort 0 1 0 1 1 (1.0) 0 1 0 0 0 Beluran 3 5 1 9 9 (1.0) 1 3 0 0 5 Keningau 12 27 8 47 47 (1.0) 1 46 0 0 0 Kinabatangan 0 1 0 1 1 (1.0) 0 1 0 0 0 Kota Belud 0 56 3 59 58 (0.98) 0 0 0 0 58 Kota Kinabalu 30 11 0 41 17 (0.42) 2 13 0 2 0 Kota Marudu 0 22 1 23 23 (1.0) 0 22 0 1 0 Kuala Penyu 0 0 0 0 0 0 0 0 0 0 Kudat 19 21 2 42 29 (0.69) 4 21 1 0 3 Kunak 4 0 0 4 1 (0.25) 0 0 0 0 1 Lahad Datu 19 27 0 46 43 (0.94) 12 21 6 2 2 Nabawan 0 107 0 107 107 (1.0) 0 46 0 22 39 Papar 7 21 0 28 28 (1.0) 0 28 0 0 0 Penampang 41 34 2 77 58 (0.75) 1 51 1 5 0 Pitas 1 0 0 1 1 (1.0) 0 1 0 0 0 Putatan 8 23 0 31 11 (0.36) 0 11 0 0 0 Ranau 0 0 0 0 0 0 0 0 0 0 Sandakan 1 0 0 1 1 (1.0) 0 1 0 0 0 Semporna 0 0 0 0 0 0 0 0 0 0 Sipitang 0 0 1 1 1 (1.0) 0 1 0 0 0 Tambunan 0 0 0 0 0 0 0 0 0 0 Tawau 96 58 4 158 158 (1.0) 57 98 3 0 0 Tenom 2 1 3 6 6 (1.0) 0 6 0 0 0 Tongod 0 1 0 1 1 (1.0) 0 0 0 1 0 Tuaran 12 21 2 35 17 (0.49) 5 12 0 0 0 Total 255 437 27 719 618 (0.86) 83 (0.13) 383 (0.62) 11 (0.02) 33 (0.05) 108 (0.17) HI = proportion of larvae‐positive houses out of all inspected houses. ‘Undetermined’ = larvae whose identification was incomplete or pending.
Chapter 3: Results for Objective 1 86
3.2. Additional analyses of dengue data
Space‐time cluster analyses were also completed for the dengue data, using SaTScan software (Kulldorff M. and Information Management Services, Inc., version 8.0), but were not ultimately included in the manuscript. SaTScan was found to be a valuable tool in its ability to detect clusters in both space and time; however, its use of ‘global’ spatial autocorrelation statistics (comparing incidence in each district to the state mean) rather than ‘local’ (where districts rates are adjusted by those of their immediate neighbours, and then compared to the state mean) resulted in a tendency to over‐estimate the size of clusters. It was also determined that performing cluster analysis at the district level was too broad a scale (and too small a sample size of districts) to yeild appropriate or reliable results. Were a finer spatial unit available to measure in Sabah (i.e. towns or localities), spatial cluster analyses might be more informative, but this was not the case with the data set available for this study.
Yet, the findings of these analyses do provide some support to those of the manuscript; in particular, showing that high incidence districts in the north‐west of the state (such as Kota Marudu) were included within annual clusters in several years (S3.4 Fig, S3.5 Table). An apparent increase in the number of clusters also occurs from 2013 onwards.
Chapter 3: Results for Objective 1 87
S3.4 Figure. Annual space‐time clusters of dengue in Sabah.
The annual number and location of space‐time dengue clusters are shown. Districts included within a space‐time cluster are those whose incidence rates were significantly above the state mean during the given time period. Details on the timing, likelihood and relative risk of each cluster are found in S3.5 Table.
Chapter 3: Results for Objective 1 88
S3.5 Table. Details of dengue space‐time cluster events in Sabah. No. of Total Radius Time frame that cluster Relative Year clusters District(s) duration LRR P value (km) occurred Risk detected* (months) 2010 1 Beluran, Nabawan, Tambunan, Ranau, Keningau 90.4 Jun 2010 to Sep 2010 4 1.6 12.8 0.0009 Sandakan, Kinabatangan, Lahad Datu, Beluran, Pitas, Kota 2011 1 178.2 Jan 2011 to Mar 2011 3 2.7 30.7 < 0.0001 Marudu, Semporna, Kudat 2012 1 Kudat, Pitas, Kota Marudu, Kota Belud, Tuaran 105.3 Jul 2012 to Dec 2012 6 6.5 183.0 < 0.0001 1 Kudat, Pitas, Kota Marudu, Kota Belud, Tuaran 105.3 Sep 2013 to Dec 2013 4 2.6 22.2 < 0.0001 2013 2 Lahad Datu, Semporna, Tawau, Kunak 61.1 Sep 2013 to Dec 2013 4 1.8 15.2 0.0002 1 Kunak, Lahad Datu, Semporna 45.8 Jun 2014 to Nov 2014 6 2.3 49.4 < 0.0001 2014 2 Tuaran, Kota Marudu, Kota Kinabalu, Ranau, Kota Belud 59.3 Dec 2014 1 2.9 34.6 < 0.0001 2015 1 Kota Kinabalu, Penampang, Putatan 7.9 Jan 2015 to Mar 2015 3 6.6 694.7 < 0.0001 1 Keningau, Tenom, Nabawan 44.5 May 2016 to Oct 2016 6 5.9 607.3 < 0.0001 2016 Kota Marudu, Kudat, Tuaran, Kota Kinabalu, Penampang, 2 104.3 Jul 2016 to Nov 2016 5 1.7 79.1 < 0.0001 Tambunan, Putatan, Kota Belud, Pitas, Ranau * The 1st and 2nd most likely clusters detected over the 7‐year period are listed here, with the duration (in months) detailed for each. Clusters were identified when the relative risk within neighbouring districts rose significantly higher than that of all other districts. Relative Risk and Log Likelihood Ratios (LLRs) are indicated for each cluster.
Chapter 3: Results for Objective 1 89
Chapter 4: Results for Objective 2
This chapter includes one manuscript:
Murphy AK, Clennon JA, Vazquez‐Prokopec G, Jansen CC, Frentiu FD, Hafner LM, Hu W, Devine GJ. Spatial and temporal patterns of Ross River virus in South East Queensland, Australia: identification of hot spots at the rural‐urban interface.
This second manuscript built upon the methodology applied in Chapter 3 to examine RRV disease patterns in urban and rural areas at a fine spatial scale, and to statistically identify the highest risk locations. To explore the potential contribution of local environments to the patterns observed, locations with statistically high or low notification rates were compared between rural and urban areas, and with the surrounding land use types. Land use type was used as a proxy indicator of the vector and host habitats present. This approach aimed to a) use a statistical approach to identify high risk areas for RRV in SEQ, b) infer potential vectors and hosts contributing to transmission in those areas, and c) provide a foundation for further investigations to incriminate specific vectors and hosts.
This chapter first describes the current demographic, spatial and temporal trends of RRV in South East Queensland. It provides an update to the literature, describing local disease patterns between 2001 and 2016, and extending upon the most recent published data from 2001. Disease patterns and risk factors were explored within 9 Local Government Areas, over the 16‐year period, at both broad and fine spatial scales. Analyses included calculation of both raw and Spatial Empirical Bayes smoothed incidence rates between rural and urban suburbs, and analysed the presence of high incidence hot spots using ‘local’ spatial autocorrelation statistics.
Hot spot patterns were visually compared to rural and urban locations and to land use types to explore potential links to environmental drivers of disease. The findings of this chapter represent an update on contemporary RRV trends in the
Chapter 4: Results for Objective 2 90
region, and report spatial trends over a broader geographical range and in more detail than has been performed in previous studies. Additionally, this chapter provides a framework for further explorations of specific associations between high RRV incidence locations and the vector and host species present, in specific land use types present, and in relation to seasonal changes in risk factors. The spatial and temporal trends identified through this work were used to inform the design of the field study described in Chapter 6.
Chapter 4: Results for Objective 2 91 Statement of Contribution of Co-Authors for Thesis by Published Paper
The Co-authors Clennon JA, Vazquez-Prokopec G, Jansen CC, Frentiu FD, Hafner LM,
Hu W, and Devine GJ of the manuscript below have certified that:
1. they meet the criteria for authorship in that they have participated in the
conception, execution, or interpretation, of at least that part of the publication
in their field of expertise;
2. they take public responsibility for their part of the publication, except for the
responsible author who accepts overall responsibility for the publication;
3. there are no other authors of the publication according to these criteria;
4. potential conflicts of interest have been disclosed to (a) granting bodies, (b) the
editor or publisher of journals or other publications, and (c) the head of the
responsible academic unit, and
5. they agree to the use of the publication in the student's thesis and its publication
on the QUT's ePrints site consistent with any limitations set by publisher
requirements.
This manuscript: Spatial and temporal patterns of Ross River virus in South East Queensland, Australia: identification of hot spots at the rural-urban interface, was submitted to BMC Infectious Diseases on 9 March, 2020 and is in the final stages of
the review process. Individual contributions of each author are noted within the
manuscript.
Principal Supervisor Confirmation:
I have sighted email or other correspondence from all Co-authors confirming their
certifying authorship.
Dr Francesca D. Frentiu QUT Verified Signature Name Signature Date
Chapter 4: Results for Objective 2 92
4.1. Spatial and temporal patterns of Ross River virus in South East Queensland, Australia: identification of hot spots at the rural‐ urban interface
Amanda K. Murphy1,2*, Julie A. Clennon3, Gonzalo Vazquez‐Prokopec4, Cassie C. Jansen5, Francesca D. Frentiu2, Louise M. Hafner2, Wenbiao Hu6, Gregor J. Devine1.
1 Mosquito Control Laboratory, QIMR Berghofer Medical Research Institute, Brisbane, Australia 2 School of Biomedical Sciences, and Institute for Health and Biomedical Innovation, Queensland University of Technology, Brisbane, Australia 3 Department of Biostatistics and Bioinformatics, Rollins School of Public Health, Emory University, Atlanta, USA. 4 Department of Environmental Sciences, Emory University, Atlanta, USA 5 Communicable Diseases Branch, Queensland Health, Herston, Australia 6 School of Public Health and Social Work, Queensland University of Technology, Brisbane, Australia
* Corresponding author, e‐mail: [email protected]
Keywords: Ross River virus, arbovirus, urban, vector‐borne, epidemic, Queensland
Chapter 4: Results for Objective 2 93
Abstract Ross River virus (RRV) is responsible for the most common vector‐borne disease of humans reported in Australia. Despite regular outbreaks, ongoing morbidity and substantial economic costs, the underlying determinants of epidemics remain unclear. Public health concern about RRV has recently increased due to rising incidence rates in Australian urban centres, along with increased circulation in Pacific Island countries. Australia experienced its largest recorded outbreak of 9,544 cases in 2015, with the majority reported from South East Queensland (SEQ). We assessed the spatial and temporal distribution of notified RRV cases, and associated epidemiological features in SEQ, from 2001‐2016. This included fine‐scale analysis of disease patterns across the suburbs of the capital city of Brisbane, and those of 8 adjacent Local Government Areas. The mean annual incidence rate for the region was 41/100,000 with a consistent seasonal peak in cases between February and May. The highest RRV incidence was in adults aged from 30‐64 years (mean incidence rate: 59/100,000), and females had higher incidence rates than males (mean incidence rates: 44/100,000 and 34/100,000, respectively). Spatial patterns of disease were heterogeneous between years, and there was a wide distribution of disease across both urban and rural areas of SEQ. Overall, the highest incidence rates were reported from predominantly rural suburbs to the north of Brisbane City, with significant hot spots located in peri‐urban suburbs where residential, agricultural and conserved natural land use types intersect. Although RRV is endemic across all of SEQ, transmission is most concentrated in areas where urban and peri‐urban environments intersect. The drivers of RRV transmission across rural‐urban landscapes should be prioritised for further investigation, including identification of specific vectors and hosts that mediate human spillover.
Chapter 4: Results for Objective 2 94
Introduction Ross River virus (RRV) is an alphavirus commonly circulating in Australia and the Western Pacific, and is responsible for the most widespread and frequently reported mosquito‐borne disease in Australia [1, 2]. An average of 5,409 RRV notifications were reported across Australia each year between 2006 and 2015 – an increase of 31% compared to the previous decade [1]. RRV is maintained in enzootic cycles between multiple species of mosquitoes, animal reservoir hosts and humans [3, 4]. In humans, symptoms of infection are similar to those of other alphaviruses, such as Barmah Forest, chikungunya and Sindbis, and may include fever, rash, fatigue and polyarthritic muscle and joint pains [5]. While not fatal, RRV disease is associated with substantial morbidity and public health impact, with symptoms including persistent pain and lethargy for weeks to months following infection [6‐8]. Because there are no specific treatments for RRV, symptoms are managed through use of analgesic and anti‐inflammatory drugs. It is estimated that notified cases represent only a small proportion of the infected population, as up to 80% of infections may be asymptomatic (9). Although a vaccine has been developed, challenges in determining commercial viability have hindered its progress to market (5, 10).
Ross River virus cases are reported across all Australian states, although Queensland typically reports around half of all annual cases (average of 48%, ranging between 24‐65% during 2001‐2016). Historically considered a rural disease, cases have increasingly been observed in metropolitan areas of Perth, Brisbane, Sydney and Melbourne since the 1990s [11‐16]. The largest recorded Australian RRV epidemic occurred from late 2014, and continued through 2015, culminating in a record annual total of 9,544 cases in 2015 (1, 17). Sixty‐five percent of these cases (6,193) were from Queensland, with 4,388 reported from the capital, Brisbane – a five‐fold increase compared with the previous 4 years (18). Subsequent large outbreaks occurred in the states of New South Wales, Victoria and Western Australia in 2017, with the national total reaching 6,928 cases (1). These coincided with reports of unexpectedly high seroprevalence rates in Australia’s neighbouring Pacific Island countries (2, 19). Although sporadic outbreaks in the Pacific Islands had been
Chapter 4: Results for Objective 2 95
documented, these recent studies revealed that RRV circulates more regularly outside of Australia than previously thought (20).
RRV prevention relies solely upon mosquito control and avoidance of bites. Development of targeted strategies for RRV management is complicated by uncertainty about the specific vector and reservoir host species that are responsible for mediating epidemics (11, 21). RRV is maintained in enzootic cycles between multiple species of mosquitoes, animal reservoir hosts and humans, but transmission pathways are poorly understood (22‐24). The virus infects several vertebrate host and mosquito vector species across different habitats and climate regions of Australia (3, 11). This includes at least 40 different mosquito species, associated with freshwater and saltwater habitats [11]. In coastal areas, the estuarine species Aedes (Ae.) vigilax (in northern Australia) and Ae. camptohynchus (in southern Australia) are considered likely vectors, while major inland vector species include freshwater‐ breeding Culex (Cx.) annulirostris, Ae. procax and the urban‐associated Ae. notoscriptus [3]. Many RRV vector species also have diverse host feeding behaviours which likely vary with environmental setting [25].
Vertebrate host species that maintain RRV circulation also vary with environmental setting. In rural areas, marsupial mammals are suspected to be a major, but not sole, reservoir [20, 21]. Hosts such as humans, livestock, rodents and birds could also play a role in virus maintenance and amplification, particularly in urban areas where marsupials are less common [20]. However, the specific determinants of transmission, including the most important mosquito vectors and reservoir hosts in different habitat types and geographic locations, remain unknown (22‐24). Factors such as climate, vegetation cover, and human behaviour likely play a role in transmission, but their relative roles are not very well defined (26‐30). This diversity in the transmission cycle complicates epidemiological investigation, prediction and control [15, 22]. Hence, to be most informative, RRV studies require regional (rather than national‐scale) approaches, adapted to specific ecological settings [31].
The South East Queensland (SEQ) region, including the capital city of Brisbane, experiences regular outbreaks and high morbidity caused by RRV. Despite this, few
Chapter 4: Results for Objective 2 96
spatial and temporal analyses of RRV trends have been conducted over the past 2 decades, and none that assess the entire region. RRV is known to have been transmitted across both urban and peri‐urban areas of Brisbane since at least the 1990’s, although the specific determinants of outbreaks remain uncertain (28, 29). Studies of two large outbreaks in Brisbane indicated a wide distribution of cases across the city (14, 17), although increased risk has also been associated with living in proximity to freshwater bushland and wetland environments (29, 32). We explored the contemporary distribution and epidemiological characteristics of RRV in SEQ between 2001 and 2016. Fine‐scale spatial and temporal trends in distribution of RRV were assessed in urban and rural areas, with the aim to provide a detailed analysis of disease trends, and explore potential links between disease patterns and transmission pathways of RRV.
Methods Ethics statement
Access and use of the human notification data for SEQ was approved by the Human Research Ethics Committees of QIMR Berghofer Medical Research Institute (QIMRB) and Queensland University of Technology (QUT) (Reference number P2238). Approval to access and use case data was obtained from the Queensland Department of Health as data custodian.
Study location
Queensland is Australia’s third most populated state, with 4.9 million inhabitants. We explored patterns of notified RRV cases in nine Local Government Areas (LGAs) of South East Queensland (SEQ): City of Brisbane, City of Ipswich, Moreton Bay Region, Redland City, Logan City, City of Gold Coast, Scenic Rim Region, Sunshine Coast Region and Shire of Noosa (Fig. 1). These 9 LGAs encompass an area of approximately 22,000 km2 and together comprise 80% of the state’s population (3.2 million), including 1.1 million in the capital city of Brisbane. The region has a sub‐ tropical climate, and diverse natural ecosystems including freshwater and estuarine
Chapter 4: Results for Objective 2 97
wetlands, Mangrove shrubland and saltmarsh, Eucalypt and Melaleuca woodlands, and rainforest (33).
Fig. 4.1. Map of Australia and South East Queensland.
RRV notification data
Daily RRV notification data were obtained from the Queensland Department of Health’s Notifiable Conditions Surveillance System (NoCS) for the 16‐year period between 1st January 2001 and 31st December 2016. Case data included: disease onset date (an estimate, based on reported onset of illness at the time of presentation to a medical clinic), age, gender, and geographical location of case residence. The residential addresses of cases were aggregated to two different geographical unit classifications used by State and Territory Local Government Departments, and described in the Australian Standard Geographical Classification (34). These were the LGA and the State Suburb Code (SSC). The broadest scale unit was the LGA, which is equivalent to a large municipal area, with a population of up to 1,100,000 people (average of 40,000); while SSC represented the finest geographical unit size, equivalent to a suburb or neighbourhood, comprising populations up to 50,000 (average 1,500). The study area included a total of 774 SSCs across 9 LGAs, of which 17 were unpopulated and excluded from the analyses. Population data for LGAs and SSCs were extracted from census data from the Australian Bureau of Statistics (ABS) for census years between 2001 and 2016 (35). Annual population figures, and population by gender and age‐group, were matched to the notification data for each LGA and SSC for the calculation of incidence rates.
Rural and urban classifications
We classified each SSC as rural or urban according to the Australian Statistical Geography Standard (ASGS), which defines urban areas of Australia (36). Urban or rural designation is based on population density and urban infrastructure criteria for each Section of State (SOS), obtained from census data. These were matched to each
Chapter 4: Results for Objective 2 98
SSC using Geographical Information System (GIS) software ArcMap 10.6 (ESRI, Redlands, CA, USA). The ASGS groups urban areas into the sub‐categories: ‘Major Urban’; a combination of Urban Centres with a total population of 100,000 or more, and ‘Other Urban’; a combination of Urban Centres with a population between 1,000 and 99,999. Rural areas also have two sub‐categories: ‘Bounded Locality’; a population centre of between 200 and 999 residents, and ‘Rural Balance’; which forms the Remainder of the State/Territory. The ABS considers both categories ‘Major Urban’ and ‘Other Urban’ as urban, while ‘Bounded Locality’ and ‘Rural Balance’ are rural. In our dataset, 292 SSCs were classified as rural and 465 as urban.
Land use data
Land use maps were obtained from the Queensland Government Land Use Mapping Program (QLUMP), available from the Queensland Spatial Catalogue (37). The QLUMP maps and assesses land use patterns and changes across the state, according to the Australian Land Use and Management (ALUM) Classification (Australian Department of Agriculture, version 8, October 2016) (38). The ALUM Classification is a detailed national standard that classifies land use types in order of increasing levels of modification of the natural landscape. The 6 classes are: 1. Conservation and Natural Environments: Land is used primarily for conservation purposes, based on the maintenance of essentially natural ecosystems already present; 2. Production from Relatively Natural Environments: Land is used mainly for primary production based on limited change to the native vegetation; 3. Production from Dryland Agriculture and Plantations: Land is used mainly for primary production, based on dryland farming systems; 4. Production from Irrigated Agriculture and Plantations: Land is used mainly for primary production, based on irrigated farming; 5. Intensive uses: Land is subject to substantial modification, generally in association with closer residential settlement, commercial or industrial uses; and 6. Water. Land use maps were imported into ArcMap 10.6 (ESRI, Redlands, CA, USA) for visualisation.
Data analysis
RRV notification epidemic features were explored over time, with monthly, annual and mean incidence rates calculated for LGAs, SSCs and the overall SEQ region. Spatial and temporal patterns of RRV notification rates were compared by age
Chapter 4: Results for Objective 2 99
and sex, and by rural versus urban classification. Differences were tested for statistical significance using the Kruskal‐Wallis or Mann‐Whitney non‐parametric tests, implemented in Statistical Package for the Social Sciences (SPSS) Statistics software (IBM New York USA; version 23), with a significance level of 0.05.
Spatial analyses included calculation of smoothed incidence rates and hot spot analysis, both performed using GeoDa software (Luc Anselin, version 1.12.1.161, September 2018). Spatial smoothing of rates were performed using Spatial Empirical Bayes (EB) smoothing technique, which corrects for outliers in raw (crude) rates to increase precision, especially where raw rates are unstable or have a high variance. In our dataset, high variances were observed due to the presence of small populations with low case numbers in some SSCs. These small population numbers can have the effect of skewing incidence rates to be extremely high. The Spatial EB technique recalculates the incidence rate for each geographical unit (SSC), applying an adjustment factor to each that is calculated relative to its raw rate and that of its immediate neighbouring geographical units (queen continguity scheme) (39). This has the effect of reducing the rates of extremely high‐rate SSCs, while also increasing the rate of low‐rate SSCs. Adjustment factors calculated for each SSC are also proportional to the population at risk, so that smaller populations will have their rates adjusted considerably, whereas rates for larger populations will change less. This is because larger populations provide greater confidence in the accuracy of the rates measured.
Hot spot analyses employed the Getis and Ord for G* local spatial autocorrelation statistic (40) for detecting hot and cold spot SSCs. Hot spot analysis was conducted using both raw and smoothed annual and mean incidence rates for SSCs, using queen contiguity neighbourhood criteria (adjusting rates by the average of immediate neighbouring SSCs in any direction). We identified and reported hot and cold spots that were significant in the raw analysis, and those that were shared between raw and smoothed analyses. Because hot/cold spots identified using mean rates were heavily influenced by the large magnitude outbreaks in 2014 and 2015, we also identified hot and cold spots that were persistently detected across at least two individual years. For this, we counted how many years an individual SSC was
Chapter 4: Results for Objective 2 100
identified as a hot/cold spot, and defined those present in at least 2/16 years as persistent. Hot spots were then mapped against urban/rural classification and land use types of SEQ. Maps of population distribution, incidence rates and hot spots were created using ArcMap 10.6 (ESRI, Redlands, CA, USA).
Results Demographic trends
For the period 2001‐2016, a total of 18,115 RRV notifications were analysed across the SEQ region. During this period, the mean annual RRV notification rate across all of SEQ was 41 cases/100,000 population. Mean annual rates were higher in females versus males, at 44/100,000 population compared to 34/100,000, respectively, though this difference was not significant (Mann‐Whitney U=84, p=0.102). For both genders, the highest incidence rates occurred in the 40‐44 and 45‐ 49 age categories (69 and 66/100,000, respectively) and the lowest in the two age categories <10 years (1 and 3/100,000, respectively) (Fig. 2). The overall trend showed a gradual increase in incidence from birth up to age 29 years (mean incidence of 17/100,000 across these age groups), peaking between ages 30‐64 years (mean 59/100,000), and dropping again ≥65 years (mean 27/100,000). This trend was consistent across all years of this study. Statistical comparisons within and between these three broad age groups indicated that rates did not differ significantly between adults from 30‐64 years (Kruskal‐Wallis H=8, p=0.239), but that this group’s rates were significantly higher than those aged ≤ 29 years (Mann‐Whitney U=1175, p<0.0001), and ≥ 65 years (U=1589, p<0.0001).
Fig. 4.2. Mean annual RRV incidence in South East Queensland by age‐group, 2001‐2016.
Chapter 4: Results for Objective 2 101
Temporal trends
RRV disease occurs in SEQ throughout all months of the year; however, a distinct seasonal pattern in the timing of annual notification peaks was shared amongst all LGAs. The peak annual notification period was typically between February and May, and this was consistent across years with higher and lower notifications (Fig. 3). These months also show the highest variability in case numbers (with a monthly average of 186 cases across SEQ, SD 246). Conversely, notifications during the winter and spring months between June‐November are generally low and relatively stable across the region (monthly average of 46 cases, SD 35). The relative magnitude of RRV outbreaks across the region were variable between years, with larger and smaller outbreaks occurring in intermittent years (the monthly temporal trend is shown in Additional file 1). Long‐term trends indicated that outbreaks generally occurred synchronously across the region, rather than initially occurring in one LGA and then spreading to another. This was also the case during largest recorded outbreak of 2015, which began earlier than usual (in late 2014) and peaked in February‐March 2015, with timing consistent across all SEQ LGAs.
Fig. 4.3. Mean monthly trend of RRV case notifications in South East Queensland, 2001‐ 2016.
Spatial trends
Spatiotemporal patterns of RRV disease varied between the 9 LGAs, with the highest case numbers in the most populated LGA, Brisbane City (5,352 total cases), and the lowest in the sparsely populated Scenic Rim Region (393 total cases) (Table 1). Annual case numbers for each LGA can be found in Additional file 2. Mean annual incidence rates across the 16‐year period varied from 30/100,000 in Gold Coast City up to 130/100,000 in Noosa Shire (Table 1), with rates in Noosa Shire being significantly higher than all other LGAs (Kruskal Wallis H=50, p<0.0001; Mann‐ Whitney U=58, p=0.007 for pairwise comparison between Noosa Shire and Scenic
Chapter 4: Results for Objective 2 102
Rim LGAs). Rates were also routinely high in the Sunshine Coast Region and sporadically high in the Scenic Rim Region. These 3 areas (Noosa Shire, Scenic Rim Region and Sunshine Coast Region) had the highest incidence rates overall, and had more rural characteristics (lower population density and proportion of urban SSCs) compared with the 4 largest cities: Brisbane, Gold Coast, Ipswich and Logan (Table 1). These 4 higher‐density cities had the lowest rates of all LGAs.
Table 4.1. Summary characteristics of each Local Government Area (LGA) of South East Queensland during the study period, 2001‐2016.
At the SSC level, spatial trends were also highly varied. The spatial trend in high‐ incidence SSCs changed from one year to the next (recent annual patterns are shown in Additional file 3), but overall the SSCs with the highest raw and smoothed incidence rates were primarily located in Noosa Shire and the Sunshine Coast Region (Fig. 4, Additional file 4). However, the highest rate of all SSCs was for Amberley in Ipswich City LGA (raw mean rate 676/100,000 and smoothed mean rate of 562/100,000). The distribution of incidence rates for rural and urban SSCs in each LGA are shown in Fig. 5. The trend across LGAs showed that SSCs with the highest incidence rates tended to have low‐mid range population densities. Mean annual RRV rates were 70 cases/100,000 (smoothed rate 85/100,000) in rural SSCs, and 44/100,000 (smoothed rate 43/100,000) in urban SSCs. Both raw and smoothed incidence rates were significantly higher in rural SSCs compared to urban SSCs (Mann‐Whitney U=58,388, p=0.001 for raw; and U=24,359, p<0.0001 for smoothed).
Fig. 4.4. Mean annual RRV incidence of State Suburb Codes (SSCs) in South East Queensland, 2001‐2016.
Chapter 4: Results for Objective 2 103
Fig. 4.5. Mean annual RRV incidence of rural and urban State Suburb Codes (SSCs), 2001‐ 2016.
Hot spot analyses
Using raw mean annual incidence rates, a total of 127/757 (17%) SSCs were identified as hot spots and 352/757 (47%) as cold spots in any individual year. These increased to 178 and 458 SSCs, respectively, using smoothed rates. A comparison of raw versus smooth analyses for annual hot and cold spots across individual years was shown in Additional file 5. In either raw or smoothed analyses, 86 hot spots and 272 cold spots were persistent (present in ≥2 years). There were 14 SSCs that were both hot and cold spots in ≥2 years that were excluded, leaving 72 persistently hot and 258 persistently cold spots. Of these, 45 hot spots and 154 cold spots were identified as persistent in both raw and smooth analyses (shown in Additional file 6). Persistent hot spots were similar to those identified using mean annual rates (n=56 mean hot spots), while mean cold spots differed (n= 47 mean cold spots). Although hot spots were geographically dispersed across all LGAs, the SSCs with the most persistent hot spots were in the Sunshine Coast Region and Noosa Shire LGAs. The same hot spots were rarely detected in consecutive years, although some were detected in multiple years, up to 7/16 years (Additional file 7). Conversely, cold spots tended to persist more in the same SSCs across several years, particularly in the Scenic Rim Region where there were very low populations and cases (Additional file 6).
A visualisation of hot spots relative to rural and urban areas of SEQ is shown in Fig. 6a. Hot spots tended to be most focused around the edges of where major urban and rural areas intersect. Of the 72 persistent hot spots detected in either raw or smoothed analyses, 35 were located in urban and 37 in rural SSCs (for the 45 hot spots shared between both analyses, 19 were urban and 26 rural). There also appeared to be diverse land use types within or adjacent to hot spot SSCs (Fig. 6b). All hot spots contained some degree of urban infrastructure, and many were located in close proximity to either dryland agriculture and plantations or to major water bodies. While some more inland hot spots were surrounded largely by conservation
Chapter 4: Results for Objective 2 104
and natural environments, these environments appeared to be most often identified as cold spots (Fig. 6, Additional file 6). Very few hot spots were located centrally within major urban areas, rather than on the edge. Only one persistent hot spot SSC was identified within Brisbane City LGA (Chelmer, identified in 2/16 years) in the raw rate analysis only. The largest cities including Brisbane, Gold Coast and Logan LGAs were more commonly dominated by cold spots rather than hot spots (Additional files 5 and 6).
Fig. 4.6. Persistent high incidence hot spots in South East Queensland, 2001‐2016.
Discussion This study sought to describe RRV spatial and temporal incidence patterns in SEQ, and to identify epidemiological trends that help elucidate the drivers of virus spillover. We found that incidence rates were highest for females, for age groups between 30‐64 years, and for residents of rural suburbs (SSCs), especially those north of Brisbane. Suburbs around the edges of major urban areas were persistent annual hot spots for RRV disease. This suggests that suburbs in rural and peri‐urban areas possess characteristics that promote circulation of RRV, possibly related to specific habitat or land use types present. The specific contributors to human infection in different environments are uncertain, and this requires further investigation. In particular, identification of RRV vectors and hosts in areas where natural and urbanised environments meet.
The demographic trends we observed were comparable with those of previous Australian studies, with the highest rates in females, and in age groups between 30 and 64 years (3, 8, 9, 11, 41). Although male to female prevalence ratios have varied slightly in previous studies, no overall gender‐related risk has been apparent (3, 8). Clinical studies report that children show fewer symptoms than adults, presumably due to age‐related differences in immune responses, while symptoms tend to persist
Chapter 4: Results for Objective 2 105
longer in adults (8, 11, 41, 42). However, seroprevalence studies have shown RRV antibody seroconversion to increase with age (43‐45) suggesting that the true incidence of infection may be higher in younger age groups than indicated by notified cases. In all likelihood, exposure to RRV is probably equivalent across all ages and genders, with higher reports in adults due to differing disease manifestations with age. Disease rates will also vary with geographical region, as the tropical northern regions of Australia have higher rates than the temperate south [1]. Because few comprehensive seroprevalence studies have been conducted in SEQ, the true age‐ related burden is unknown.
There was a strikingly consistent seasonal trend in outbreaks across SEQ, peaking annually between the months of February and May. This coincides with periods of relatively high temperature and rainfall from late austral summer to early autumn, when SEQ’s average daily temperatures are 20‐24◦C (46). Although climate alone does not predict outbreak occurrence, it influences vector and wildlife host species’ abundance (47‐50). Favourable temperature conditions, rainfall, high tides and low‐level flooding have all been associated with elevated RRV risk in previous studies (26‐28, 51‐53). Temperature also impacts viral replication, with the ideal temperature for RRV transmission at 26.4 degrees Celsius (transmission range of 17 to 31.5◦C) (54). However, variations in weather patterns and vector‐host ecology mean that climate‐based predictions are only valid locally or, at best, regionally, rather than nationally (55, 56). Hence, although suitable weather conditions are a requisite precursor to outbreaks, outbreak occurrence ultimately depends on availability of, and interactions between, sufficient competent vectors and susceptible hosts (57, 58).
The variation in annual spatial trends observed suggests that conditions supporting transmission occur sporadically in particular SSCs, and may change from one year to the next. This might be due to local climatic and environmental variations which influence vector and host abundance in both freshwater and saltwater habitats (23, 27). Freshwater vectors Cx. annulirostris, Ae. notoscriptus, Ae. procax, and Ae. vittiger as well as saltwater vectors Ae. vigilax, Cx. sitiens and Verallina funerea were associated with large outbreaks in Brisbane and the Sunshine Coast Region during the
Chapter 4: Results for Objective 2 106
1990s [14, 59]. The most recent outbreaks of 2014‐2015 were linked to increased abundance of the freshwater vectors Cx. annulirostris and Ae. procax in Brisbane following high rainfall [17]. Many of the implicated vectors share similar or overlapping habitat types and have broad host‐feeding behaviours [25]. Hence, the relative contribution of different vectors is to human outbreaks is challenging to disentangle. It is possible that multiple vectors in different habitats contribute to varying degrees, at different times [60, 61].
Similarly, a number of different hosts that maintain RRV circulation across SEQ could contribute to epidemics. Although few studies have investigated the role of specific wildlife hosts in human RRV outbreaks, opportunistic serosurveys of wildlife together with a handful of experimental infection studies have generated some hypotheses (21). Potential hosts theorised to contribute to RRV transmission include birds, small mammals and marsupials (including rodents, possums, flying foxes) in urban areas; and larger mammals and marsupial macropods (such as horses and cattle, kangaroos and wallabies) in peri‐urban and rural areas (15, 20, 21, 57). However, current evidence identifying important RRV hosts is limited, and broader investigations of the transmission potential of wildlife are much needed (57). In the absence of these, it can be assumed from RRV’s wide geographic and habitat range that there is flexibility in both vectors and hosts. In SEQ, the seasonal composition of vectors and hosts in peri‐urban habitats, especially those in proximity to hot spot suburbs, should be a particular focus for future RRV transmission studies.
We identified both high incidence rates and the most persistent hot spots overall in Noosa Shire and Sunshine Coast Region, in which there are low‐medium human population densities and diverse land use types present. The specific drivers of high rates of RRV in these LGAs are uncertain, but could relate to the proximity of peri‐urban human populations to rural vector and wildlife habitats. Interactions between humans, vectors and wildlife in or near particular land use types in peri‐ urban areas could create a ‘perfect storm’ of factors supporting RRV transmission. Human‐modified and fragmented landscapes are known to influence the risk of vector borne diseases, either positively or negatively, through altering ecological relationships between wildlife, vectors and humans. (62‐64). Land use changes such
Chapter 4: Results for Objective 2 107
as deforestation and agricultural development have been linked to increased risk of West Nile virus and malaria infection (65, 66), and have been linked to arboviral disease risk in Australia (67, 68). While our findings do not confirm a link between land use and RRV risk, they do suggest this could be worthy of investigation. Given RRV’s expansion into urban and outer metropolitan areas of Australia in recent years, it is conceivable that urban expansion and alteration of wildlife habitats may have implications for RRV risk.
The absence of hot spots, and concentration of cold spots, in the Scenic Rim LGA suggests that it lacks sufficient human, wildlife host and vector populations to maintain persistent outbreaks. This is despite the Scenic Rim having a large proportion of natural conservation and irrigated agricultural areas, which could theoretically support mosquito and wildlife habitats. This might be explained by the low human population in this LGA, which results in sporadically high but inconsistent incidence rates, and unstable spatial patterns of disease (both hot and cold spots in the same suburbs in different years). This pattern could potentially change if human populations in the Scenic Rim were to increase. Again, analyses of different land use types, their association with specific vector and host habitats, and RRV risk would assist understanding of how and where these factors inter‐relate.
This study is the first to describe long‐term epidemiological trends of RRV across SEQ. We report RRV disease patterns at both broad (LGA) and fine (SSC) scales, and identify characteristics associated with higher RRV risk that can inform future investigations. However, the study was limited by its reliance on routinely collected public health data, and the associated challenges with passive disease reporting. In Queensland, notification processes for RRV do not include individual case interview, nor information on the timing and location of RRV exposure, which is often unknown. We used the case’s reported onset date of symptoms and residential address as a proxy for this. While we expect that many residents will be bitten and infected in their home suburb, this will not be true for all, and there is no way to correct for this. Nevertheless, given RRV’s high incidence and wide geographical range across SEQ, using the case residence seems a reasonable proxy for location of infection. Socioecological factors not accounted for in our study may also have influenced the
Chapter 4: Results for Objective 2 108
demographic trends we observed. For instance, healthcare‐seeking practices likely differ between genders and age‐groups, and exposure to mosquitoes through occupational or leisure activities could also differ between demographic groups. The impact of socioeconomic factors on infection risk has also varies by geographic region (22, 23, 28). However, within our study region, previous studies indicate that socioeconomic variation is unlikely to have had a significant impact on our results (28, 32).
Conclusions
Overall, our findings contribute to understanding of RRV disease patterns and public health risk in SEQ. To further progress understanding of RRV transmission risk and improve future disease prediction and prevention, greater understanding of seasonal variation in distribution and abundance of potential vectors and hosts is essential. For SEQ, this is especially important in urban fringe areas, and areas undergoing urban expansion. Peri‐urban suburbs at the rural‐urban interface, especially where different land use types intersect, could be most capable of supporting adequate densities of vectors, hosts and humans to allow persistent transmission. We recommend that hot spots identified in these ‘edge’ locations be targeted for further investigation of RRV transmission pathways. Clarifying the relative importance of specific contributors to RRV epidemics is a priority for developing targeted disease prevention strategies – and may have flow‐on benefits for prevention of other endemic and imported mosquito‐borne infections in Australia.
Chapter 4: Results for Objective 2 109
Acknowledgements
The authors would like to thank the Queensland Department of Health, Australia for making Ross River virus notification data available. We are grateful for additional assistance and input provided by Dr Fiona May, and Dr Jonathan Darbro of Metro North Public Health Unit, Dr Eloise Skinner of Griffith University, and the knowledgeable staff from the mosquito management teams the local government councils of SEQ. In particular, we would like to thank Dr Martin Shivas, Mr Michael Onn and Mr Mark Call for their helpful advice and support.
Authors’ Contributions
AM, WH, and GD conceived the project. AM carried out the analyses, with input from JC, GVP, CJ, WH and GD. AM drafted the paper. FF and LH aided interpretation. All authors reviewed and revised the final manuscript and agreed to its submission.
Chapter 4: Results for Objective 2 110
References 1. National Notifiable Diseases Surveillance System: Commonwealth of Australia; Australian Government. [Available from: http://www9.health.gov.au/cda/source/cda‐index.cfm. 2. Lau C, Aubry M, Musso D, Teissier A, Paulous S, Despres P, et al. New evidence for endemic circulation of Ross River virus in the Pacific Islands and the potential for emergence. International journal of infectious diseases : IJID : official publication of the International Society for Infectious Diseases. 2017;57:73‐6. 3. Harley D, Sleigh A, Ritchie S. Ross River virus transmission, infection, and disease: a cross‐disciplinary review. Clin Microbiol Rev. 2001;14(4):909‐32, table of contents. 4. Ratnayake JTB. PhD Thesis: The Valuation of Social and Economic Costs of Mosquito‐Transmitted Ross River Virus: Griffith University; 2006. 5. Woodruff R, Bambrick H. Climate change impacts on the burden of Ross River virus disease. National Library of Australia.: Commonwealth of Australia; Garnaut Climate Change Review. June 2008. 6. Condon RJ, Rouse IL. Acute symptoms and sequelae of Ross River virus infection in South‐Western Australia: a follow‐up study. Clinical and diagnostic virology. 1995;3(3):273‐84. 7. Suhrbier A, La Linn M. Clinical and pathologic aspects of arthritis due to Ross River virus and other alphaviruses. Curr Opin Rheumatol. 2004;16(4):374‐9. 8. Selden SM, Cameron AS. Changing epidemiology of Ross River virus disease in South Australia. Med J Aust. 1996;165(6):313‐7. 9. Aaskov JG, Ross P, Davies CE, Innis MD, Guard RW, Stallman ND, et al. Epidemic polyarthritis in northeastern Australia, 1978‐1979. Med J Aust. 1981;2(1):17‐9. 10. Tomerini DM, Dale PE, Sipe N. Does mosquito control have an effect on mosquito‐borne disease? The case of Ross River virus disease and mosquito management in Queensland, Australia. J Am Mosq Control Assoc. 2011;27(1):39‐44. 11. Russell RC. Ross River virus: ecology and distribution. Annual review of entomology. 2002;47:1‐31. 12. Lindsay M, Oliveira N, Jasinska E, Johansen C, Harrington S, Wright AE, et al. An outbreak of Ross River virus disease in Southwestern Australia. Emerging infectious diseases. 1996;2(2):117‐20. 13. Brokenshire T, Symonds D, Reynolds R, Doggett S, Geary M, Russell R. A cluster of locally‐acquired Ross River virus infection in outer western Sydney. N S W Public Health Bull. 2000;11(7):132‐4. 14. Ritchie SA, Fanning ID, Phillips DA, Standfast HA, McGinn D, Kay BH. Ross River virus in mosquitoes (Diptera:Culicidae) during the 1994 epidemic around Brisbane, Australia. J Med Entomol. 1997;34(2):156‐9. 15. Claflin SB, Webb CE. Ross River Virus: Many Vectors and Unusual Hosts Make for an Unpredictable Pathogen. PLoS pathogens. 2015;11(9):e1005070. 16. Tong S, Bi P, Hayes J, Donald K, Mackenzie J. Geographic variation of notified Ross River virus infections in Queensland, Australia, 1985‐1996. The American journal of tropical medicine and hygiene. 2001;65(3):171‐6.
Chapter 4: Results for Objective 2 111
17. Jansen CC, Shivas MA, May FJ, Pyke AT, Onn MB, Lodo K, et al. Epidemiologic, Entomologic, and Virologic Factors of the 2014–15 Ross River Virus Outbreak, Queensland, Australia. Emerging infectious diseases. 2019;25(12). 18. Notifiable Conditions Reports: Queensland Government; Queensland Department of Health. [Available from: https://www.health.qld.gov.au/clinical‐ practice/guidelines‐procedures/diseases‐ infection/surveillance/reports/notifiable/annual/default.asp. 19. Aubry M, Finke J, Teissier A, Roche C, Broult J, Paulous S, et al. Silent Circulation of Ross River Virus in French Polynesia. International journal of infectious diseases : IJID : official publication of the International Society for Infectious Diseases. 2015;37:19‐24. 20. Flies EJ, Lau CL, Carver S, Weinstein P. Another Emerging Mosquito‐Borne Disease? Endemic Ross River Virus Transmission in the Absence of Marsupial Reservoirs. BioScience. 2018;68(4):288‐93. 21. Stephenson EB, Peel AJ, Reid SA, Jansen CC, McCallum H. The non‐human reservoirs of Ross River virus: a systematic review of the evidence. Parasites & vectors. 2018;11(1):188. 22. Tong S, Dale P, Nicholls N, Mackenzie JS, Wolff R, McMichael AJ. Climate variability, social and environmental factors, and ross river virus transmission: research development and future research needs. Environ Health Perspect. 2008;116(12):1591‐7. 23. Kelly‐Hope LA, Purdie DM, Kay BH. Ross River virus disease in Australia, 1886‐ 1998, with analysis of risk factors associated with outbreaks. J Med Entomol. 2004;41(2):133‐50. 24. Jardine A, Neville PJ, Lindsay MD. Proximity to mosquito breeding habitat and Ross River virus risk in the Peel region of Western Australia. Vector borne and zoonotic diseases. 2015;15(2):141‐6. 25. Stephenson EB, Murphy AK, Jansen CC, Peel AJ, McCallum H. Interpreting mosquito feeding patterns in Australia through an ecological lens: an analysis of blood meal studies. Parasites & vectors. 2019;12(1):156. 26. Hu W, Nicholls N, Lindsay M, Dale P, McMichael AJ, Mackenzie JS, et al. Development of a predictive model for ross river virus disease in Brisbane, Australia. The American journal of tropical medicine and hygiene. 2004;71(2):129‐37. 27. Tong S, Hu W, Nicholls N, Dale P, MacKenzie JS, Patz J, et al. Climatic, high tide and vector variables and the transmission of Ross River virus. Intern Med J. 2005;35(11):677‐80. 28. Hu W, Clements A, Williams G, Tong S, Mengersen K. Bayesian spatiotemporal analysis of socio‐ecologic drivers of Ross River virus transmission in Queensland, Australia. The American journal of tropical medicine and hygiene. 2010;83(3):722‐8. 29. Muhar A, Dale PE, Thalib L, Arito E. The spatial distribution of ross river virus infections in Brisbane: Significance of residential location and relationships with vegetation types. Environ Health Prev Med. 2000;4(4):184‐9. 30. Ng V, Dear K, Harley D, McMichael A. Analysis and prediction of Ross River virus transmission in New South Wales, Australia. Vector borne and zoonotic diseases. 2014;14(6):422‐38.
Chapter 4: Results for Objective 2 112
31. Flies EJ, Weinstein P, Anderson SJ, Koolhof I, Foufopoulos J, Williams CR. Ross River Virus and the Necessity of Multiscale, Eco‐epidemiological Analyses. J Infect Dis. 2018;217(5):807‐15. 32. Hu W, Tong S, Mengersen K, Oldenburg B. Exploratory spatial analysis of social and environmental factors associated with the incidence of Ross River virus in Brisbane, Australia. The American journal of tropical medicine and hygiene. 2007;76(5):814‐9. 33. Regional ecosystem descriptions Queensland Government. [updated April 16, 2019. Available from: https://apps.des.qld.gov.au/regional‐ ecosystems/list/?bioregion=12. 34. Australian Standard Geographical Classification (AGSC) Canberra, AustraliaAustralian Bureau of Statistics; October 5, 2011. [Available from: http://www.abs.gov.au/AUSSTATS/[email protected]/Previousproducts/1216.0Content s11999?opendocument&tabname=Summary&prodno=1216.0&issue=1999&nu m=&view=. 35. Australian census data by geography: Commonwealth of Australia; Australian Bureau of Statistics. [Available from: https://www.abs.gov.au/websitedbs/D3310114.nsf/Home/Census?OpenDocu ment&ref=topBar. 36. Australian Statistical Geography Standard (ASGS): Volume 3 ‐ Non ABS Structures. Australian Bureau of Statistics. 37. Queensland Spatial Catalogue ‐ QSpatial: Queensland Government; Queensland State Department of Natural Resources, Mines and Energy. [updated June 4, 2019. Available from: http://qldspatial.information.qld.gov.au/catalogue/custom/index.page. 38. Australian Land Use and Management Classification Version 8. Canberra, AustraliaDepartment of Agriculture and Water Resources, Australian Bureau of Agricultural and Resource Economics and Sciences. 39. Anselin L, Lozano‐Gracia N, Koschinky J. Rate Transformations and Smoothing. Technical Report., Spatial Analysis Laboratory DoG; 2006. 40. Ord JK, Getis A. Local Spatial Autocorrelation Statistics: Distributional Issues and Application. Geographical Analysis. 1995;27(4). 41. Hawkes RA, Boughton CR, Naim HM, Stallman ND. A major outbreak of epidemic polyarthritis in New South Wales during the summer of 1983/1984. Med J Aust. 1985;143(8):330‐3. 42. Mackenzie JS, Smith DW. Mosquito‐borne viruses and epidemic polyarthritis. Med J Aust. 1996;164(2):90‐3. 43. Boughton CR, Hawkes RA, Naim HM, Wild J, Chapman B. Arbovirus infections in humans in New South Wales. Seroepidemiology of the alphavirus group of togaviruses. Med J Aust. 1984;141(11):700‐4. 44. Hawkes RA, Pamplin J, Boughton CR, Naim HM. Arbovirus infections of humans in high‐risk areas of south‐eastern Australia: a continuing study. Med J Aust. 1993;159(3):159‐62. 45. Doherty RL. Surveys of haemagglutination‐inhibiting antibody to arboviruses in Aborigines and other population groups in Northern and Eastern Australia, 1966‐1971. Trans R Soc Trop Med Hyg. 1973;67(2):197‐205.
Chapter 4: Results for Objective 2 113
46. Climate Data Online: Commonwealth of Australia; Australian Bureau of Meteorology. [Available from: http://www.bom.gov.au/jsp/ncc/climate_averages/temperature/index.jsp. 47. Williams CR, Fricker SR, Kokkinn MJ. Environmental and entomological factors determining Ross River virus activity in the River Murray Valley of South Australia. Aust N Z J Public Health. 2009;33(3):284‐8. 48. Jacups SP, Whelan PI, Markey PG, Cleland SJ, Williamson GJ, Currie BJ. Predictive indicators for Ross River virus infection in the Darwin area of tropical northern Australia, using long‐term mosquito trapping data. Tropical medicine & international health : TM & IH. 2008;13(7):943‐52. 49. Ritchie EG, Bolitho EE. Australia's savanna herbivores: bioclimatic distributions and an assessment of the potential impact of regional climate change. Physiol Biochem Zool. 2008;81(6):880‐90. 50. Martin GA, Yanez‐Arenas C, Roberts BJ, Chen C, Plowright RK, Webb RJ, et al. Climatic suitability influences species specific abundance patterns of Australian flying foxes and risk of Hendra virus spillover. One Health. 2016;2:115‐21. 51. Tall JA, Gatton ML. Flooding and Arboviral Disease: Predicting Ross River Virus Disease Outbreaks Across Inland Regions of South‐Eastern Australia. J Med Entomol. 2019. 52. Woodruff RE, Guest CS, Garner MG, Becker N, Lindesay J, Carvan T, et al. Predicting Ross River virus epidemics from regional weather data. Epidemiology. 2002;13(4):384‐93. 53. Tong S, Bi P, Donald K, McMichael AJ. Climate variability and Ross River virus transmission. J Epidemiol Community Health. 2002;56(8):617‐21. 54. Shocket MS, Ryan SJ, Mordecai EA. Temperature explains broad patterns of Ross River virus transmission. Elife. 2018;7. 55. Jacups SP, Whelan PI, Currie BJ. Ross River virus and Barmah Forest virus infections: a review of history, ecology, and predictive models, with implications for tropical northern Australia. Vector borne and zoonotic diseases. 2008;8(2):283‐97. 56. Kelly‐Hope LA, Purdie DM, Kay BH. El Nino Southern Oscillation and Ross River virus outbreaks in Australia. Vector borne and zoonotic diseases. 2004;4(3):210‐ 3. 57. Carver S, Bestall A, Jardine A, Ostfeld RS. Influence of hosts on the ecology of arboviral transmission: potential mechanisms influencing dengue, Murray Valley encephalitis, and Ross River virus in Australia. Vector borne and zoonotic diseases. 2009;9(1):51‐64. 58. Koolhof IS, Carver S. Epidemic host community contribution to mosquito‐borne disease transmission: Ross River virus. Epidemiology and infection. 2017;145(4):656‐66. 59. Ryan PA, Do KA, Kay BH. Spatial and temporal analysis of Ross River virus disease patterns at Maroochy Shire, Australia: association between human morbidity and mosquito (Diptera: Culicidae) abundance. J Med Entomol. 1999;36(4):515‐21. 60. Hu W, Mengersen K, Dale P, Tong S. Difference in mosquito species (Diptera: Culicidae) and the transmission of Ross River virus between coastline and
Chapter 4: Results for Objective 2 114
inland areas in Brisbane, Australia. Environmental entomology. 2010;39(1):88‐ 97. 61. Hu W, Tong S, Mengersen K, Oldenburg B, Dale P. Mosquito species (Diptera: Culicidae) and the transmission of Ross River virus in Brisbane, Australia. J Med Entomol. 2006;43(2):375‐81. 62. Steiger DB, Ritchie SA, Laurance SG. Land Use Influences Mosquito Communities and Disease Risk on Remote Tropical Islands: A Case Study Using a Novel Sampling Technique. The American journal of tropical medicine and hygiene. 2016;94(2):314‐21. 63. Kilpatrick AM, Randolph SE. Drivers, dynamics, and control of emerging vector‐ borne zoonotic diseases. Lancet. 2012;380(9857):1946‐55. 64. Gottdenker NL, Streicker DG, Faust CL, Carroll CR. Anthropogenic land use change and infectious diseases: a review of the evidence. Ecohealth. 2014;11(4):619‐32. 65. Gomez A, Kilpatrick AM, Kramer LD, Dupuis AP, 2nd, Maffei JG, Goetz SJ, et al. Land use and west nile virus seroprevalence in wild mammals. Emerging infectious diseases. 2008;14(6):962‐5. 66. Fornace KM, Abidin TR, Alexander N, Brock P, Grigg MJ, Murphy A, et al. Association between Landscape Factors and Spatial Patterns of Plasmodium knowlesi Infections in Sabah, Malaysia. Emerging infectious diseases. 2016;22(2):201‐8. 67. Meyer Steiger DB, Ritchie SA, Laurance SG. Mosquito communities and disease risk influenced by land use change and seasonality in the Australian tropics. Parasites & vectors. 2016;9(1):387. 68. Walsh MG. Ecological and life history traits are associated with Ross River virus infection among sylvatic mammals in Australia. BMC Ecol. 2019;19(1):2.
Chapter 4: Results for Objective 2 115
Tables and Figures
Chapter 4: Results for Objective 2 116
Figure 4.1. Map of Australia and South East Queensland.
On left, the 8 Australian states are shown including three largest Australian cities Sydney, Melbourne and Brisbane, marked by black dots. Inset, South East Queensland, and the 9 local government areas (LGAs) included in the study, each sub‐divided into smaller State Suburb Code (SSC) units. The population distribution for SSCs is indicated by graduated shading.
Chapter 4: Results for Objective 2 117
Figure 4.2. Mean annual RRV incidence rates by age‐group in South East Queensland, 2001‐2016.
Age‐adjusted mean annual incidence rates are grouped into 18 age groups of SEQ, between 0 and 95 years. Mean rates per group are indicated by X, and the median by the horizontal line across each box. Whiskers indicate the estimated minimum (lower whisker) and maximum (upper whisker) incidence values per age group (equal to the 1st and 3rd quartile ‐/+ 1.5 x the inter‐quartile range, respectively). Outlying points indicate extreme values reported during large outbreak years, e.g. during the outbreaks of 2014 and 2015.
Chapter 4: Results for Objective 2 118
Figure 4.3. Mean monthly trend of RRV case notifications in South East Queensland, 2001‐ 2016.
Mean monthly case numbers across the 16‐year period are shown, with extreme case report values from large outbreak years indicated as outliers. Mean rates per group are indicated by X, and median by the horizontal line crossing each box. Whiskers indicate the estimated maximum and minimum case numbers per month across the 16 years (equal to the 1st and 3rd quartile ‐/+ 1.5 x the inter‐quartile range, respectively). *Total cases reported in February 2015 (1,357) were the highest ever reported in a single month.
Chapter 4: Results for Objective 2 119
Table 4.1. Summary characteristics of each Local Government Area (LGA) of South East Queensland during the study period, 2001‐2016.
Average Population Proportion Mean annual Total cases LGA name* population density of urban incidence rate (2001‐2016) (2001‐2016) (per km2) SSCs (per 100,000)
Brisbane City 1,023,663 762 5,352 94.2% 33 Gold Coast City 475,169 356 2,252 74.1% 30 Ipswich City 157,963 146 1,044 58.0% 41 Logan City 271,082 283 1,456 81.7% 34 Moreton Bay 357,532 175 3,097 58.5% 54 Region Noosa Shire 41,008 47 850 29.2% 130 Redland City 134,328 250 991 63.6% 46 Scenic Rim 35,443 8 393 2.8% 69 Region Sunshine Coast 256,982 114 2,680 50.4% 65 Region Total 2,753,170 188 18,115 61.0% 41 * See Fig. 4.1 for location of each LGA within the study area.
Chapter 4: Results for Objective 2 120
Figure 4.4. Mean annual RRV incidence of State Suburb Codes (SSCs) in South East Queensland, 2001‐2016. a) Mean annual incidence rate, and b) smoothed mean annual incidence rate for SSCs of South East Queensland over the 16‐year study period. Seventeen unpopulated SSCs are indicated by the white (zero case) locations in b).
Chapter 4: Results for Objective 2 121
Figure 4.5. Mean annual RRV incidence of rural and urban State Suburb Codes (SSCs), 2001‐2016. a) raw incidence rates, and b) smoothed incidence rates for the 757 SSCs of South East Queensland, shown according to the 9 Local Government Areas (LGAs) they are located within, and all SSCs combined. Rural and urban SSCs are indicated by blue and red circles, respectively. Two outlying SSCs (in Ipswich City and Noosa Shire) had incidence values beyond the scale shown here, but are listed in S4.4 Table. See Fig. 4.1 for location of each LGA within the study area.
Chapter 4: Results for Objective 2 122
Figure 4.6. Persistent high incidence hot spots in South East Queensland, 2001‐2016.
Persistent hot spots identified for State Suburb Codes (SSCs) are shown relative to a) urban and rural areas of South East Queensland; and b) different land use types of South East Queensland. Points indicate the centroid of each hot spot SSC. Hot spots detected using raw incidence rates only (yellow points; n=27 SSCs) and those detected in both raw and smoothed incidence analyses (black points; n=45 SSCs) are indicated. Note: in b) ‘Intensive uses’ refers to residential areas and urban infrastructure. See methods for further description.
Chapter 4: Results for Objective 2 123
Supporting Information
S4.1 Figure. Monthly trend of RRV notifications in South East Queensland, 2001‐ 2016.
S4.2 Table. Summary of annual case counts for each Local Government Area (LGA).
S4.3 Figure. Annual RRV incidence in South East Queensland: 2013‐2016.
S4.4 Table. Locations with the highest overall rates across all years, 2001‐2016.
S4.5 Figure. Annual hot and cold spots for RRV incidence in South East Queensland: 2013‐2016.
S4.6 Figure. Persistent and mean RRV hot spots in South East Queensland: 2001‐ 2016.
S4.7 Table. Summary of 45 persistent hot spots identified in both raw and smoothed incidence analyses from 2001‐2016.
Chapter 4: Results for Objective 2 124
S4.1 Figure. Monthly trend of RRV notifications in South East Queensland, 2001‐2016.
Monthly case notifications are shown for each of the 9 Local Government Areas (LGAs) in the study area. A major flooding event that occurred in the region in early 2011 likely reduced case numbers of that year by inundating vector breeding sites with fast‐flowing water. The largest ever recorded peak in monthly cases occurred for all LGAs during February and March 2015.
Chapter 4: Results for Objective 2 125
S4.2 Table. Summary of annual case counts for each Local Government Area (LGA). Grand LGA name 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Total Brisbane City 325 39 363 342 74 454 239 418 224 299 168 292 231 272 1,477 135 5,352 Gold Coast City 106 21 153 99 37 153 125 138 121 122 61 102 107 164 636 107 2,252 Ipswich City 28 11 52 92 6 80 47 76 54 66 45 44 41 65 300 37 1,044 Logan City 95 11 145 99 24 117 68 111 60 60 30 59 48 59 424 46 1,456 Moreton Bay Region 150 26 289 176 73 212 190 229 160 186 112 175 160 182 672 105 3,097 Noosa Shire 29 12 161 48 26 35 36 63 46 68 15 44 34 85 103 45 850 Redland City 97 13 69 33 17 49 70 64 37 61 44 51 40 47 255 44 991 Scenic Rim Region 7 4 22 41 5 42 13 24 12 31 14 15 14 10 122 17 393 Sunshine Coast Region 98 41 342 111 64 123 143 205 195 234 89 123 109 236 399 168 2,680 Total 935 178 1,596 1,041 326 1,265 931 1,328 909 1,127 578 905 784 1,120 4,388 704 18,115
Chapter 4: Results for Objective 2 126
S4.3 Figure. Annual RRV incidence in South East Queensland: 2013‐2016.
Annual incidence patterns are shown for State Suburb Codes (SSCs) within each of the 9 Local Government Areas in the years before, during and after the largest recorded epidemic in 2015.
Chapter 4: Results for Objective 2 127
S4.4 Table. Locations with the highest overall rates across all years, 2001‐2016.
Population Mean smoothed Urban/rural Average Total cases Mean raw annual LGA name SSC name density annual incidence classification1 population2 (2001‐2016) incidence rate* (persons/km2) rate* Ipswich City Amberley Rural Balance 222 8 24 676 562 Noosa Shire Boreen Point Rural Balance 287 55 19 413 363 Noosa Shire Noosa North Shore Rural Balance 157 1 9 359 176 Scenic Rim Region Woolooman Rural Balance 20 1 1 310 67 Gold Coast City Gilberton Rural Balance 22 2 1 285 61 Noosa Shire Cooroy Rural Balance 3,323 121 142 267 263 Sunshine Coast Region Kunda Park Major Urban 24 9 1 264 71 Scenic Rim Region Kents Lagoon Rural Balance 49 5 2 255 147 Noosa Shire Cooran Rural Balance 1,423 38 56 246 235 Sunshine Coast Region Eumundi Rural Balance 1,947 83 75 241 234 Noosa Shire Pomona Rural Balance 2,551 48 98 240 237 Scenic Rim Region Mount Edwards Rural Balance 55 4 2 226 125 Sunshine Coast Region Yandina Other Urban 2,078 136 72 217 207 Sunshine Coast Region Conondale Rural Balance 752 4 26 216 202 Sunshine Coast Region Elaman Creek Rural Balance 60 5 2 210 138 Ipswich City Blacksoil Rural Balance 91 48 3 206 74 Redland City Lamb Island Bounded Locality 379 273 12 198 0
Chapter 4: Results for Objective 2 128
Population Mean smoothed Urban/rural Average Total cases Mean raw annual LGA name SSC name density annual incidence classification1 population2 (2001‐2016) incidence rate* (persons/km2) rate* Moreton Bay Region Samsonvale Rural Balance 517 15 16 193 135 Sunshine Coast Region Kenilworth Rural Balance 489 3 15 192 173 Redland City Point Lookout Rural Balance 625 139 19 190 177 Scenic Rim Region Mutdapilly Rural Balance 270 6 8 185 139 Scenic Rim Region Moogerah Rural Balance 205 1 6 183 123 Sunshine Coast Region Peachester Rural Balance 1,189 26 34 179 138 Gold Coast City Norwell Rural Balance 175 7 5 178 98 Noosa Shire Kin Kin Rural Balance 670 7 19 177 183 Brisbane City Fairfield Major Urban 859 725 24 175 139 Scenic Rim Region Roadvale Rural Balance 251 14 7 175 99 Scenic Rim Region Harrisville Rural Balance 537 17 15 174 132 Brisbane City Pinkenba Major Urban 323 23 9 174 0 Sunshine Coast Region Doonan Other Urban 3,032 94 84 173 168 Sunshine Coast Region Coochin Creek Rural Balance 74 1 2 170 95 Scenic Rim Region Kalbar Rural Balance 958 30 26 170 155 Scenic Rim Region Canungra Rural Balance 1,077 36 29 168 153 Scenic Rim Region Innisplain Rural Balance 75 3 2 168 57 Ipswich City Purga Rural Balance 505 9 13 161 154 Scenic Rim Region Mount Walker Rural Balance 117 4 3 160 93 Gold Coast City Jacobs Well Other Urban 1,612 119 41 159 149
Chapter 4: Results for Objective 2 129
Population Mean smoothed Urban/rural Average Total cases Mean raw annual LGA name SSC name density annual incidence classification1 population2 (2001‐2016) incidence rate* (persons/km2) rate* Brisbane City Kooringal Rural Balance 39 249 1 158 72 Moreton Bay Region Closeburn Rural Balance 521 44 13 156 106 Scenic Rim Region Moorang Rural Balance 40 1 1 155 88 Scenic Rim Region Coleyville Rural Balance 162 4 4 154 135 Moreton Bay Region Laceys Creek Rural Balance 244 3 6 154 109 Redland City Russell Island Other Urban 2,486 142 61 153 0 Sunshine Coast Region Eudlo Rural Balance 979 49 24 153 113 Moreton Bay Region Mount Pleasant Rural Balance 291 8 7 150 116 Sunshine Coast Region Verrierdale Rural Balance 679 23 16 147 139 Moreton Bay Region Ocean View Rural Balance 820 21 19 145 99 1 Based on ASGS classification of urban and rural, available from the Australian Bureau of Statistics. 2 Based on census data, available from the Australian Bureau of Statistics. * per 100,000 population
Chapter 4: Results for Objective 2 130
S4.5 Figure. Annual hot and cold spots for RRV incidence in South East Queensland: 2013‐2016.
Significant high‐ and low‐incidence (hot and cold) spots identified through two different analysis techniques are overlaid: local G* analysis of raw (crude) annual incidence rates for State Suburb Codes (SSCs), and smoothed annual rates for SSCs (Empirical Bayes Spatial smoothing technique). Disagreement occurred where an SSC was hot in one analysis and cold in the other, or vice versa.
Chapter 4: Results for Objective 2 131
S4.6 Figure. Persistent and mean RRV hot spots in South East Queensland: 2001‐2016.
Significant high‐ and low‐incidence (hot and cold) spots shared between raw and smoothed incidence rate analyses are shown by State Suburb Code (SSC): a) 45 persistent hot and 154 persistent cold spots (present in ≥ 2 years) present in both raw and smoothed analyses; and b) 56 mean hot and 47 mean cold spots present in both raw and smoothed analyses. In a) SSC colours are graduated according to the number of years identified as a hot/cold spot, including 14 additional SSCs that were both hot and cold in ≥ 2/16 years.
Chapter 4: Results for Objective 2 132
S4.7 Table. Summary of 45 persistent hot spots identified in both raw and smoothed incidence analyses from 2001‐2016.
Total Rural/ Total hot LGA name SSC name urban 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 cases spot category years Sunshine Other Doonan 84 0 1 1 0 1 0 0 0 1 1 0 1 0 1 0 0 7 Coast Region Urban Major Noosa Shire Tewantin 167 0 0 1 0 1 0 0 1 0 1 0 0 0 1 0 0 5 Urban Rural Noosa Shire Cooroy 142 0 0 1 0 0 0 0 0 1 0 0 1 1 1 0 0 5 Balance Major Noosa Shire Noosaville 98 0 0 1 0 1 0 0 0 1 1 0 0 0 1 0 0 5 Urban Sunshine Rural Eumundi 75 0 0 1 0 1 0 0 1 0 0 0 0 0 1 0 1 5 Coast Region Balance Rural Noosa Shire Cooroibah 40 0 0 1 0 1 0 0 0 0 1 0 0 1 1 0 0 5 Balance Rural Noosa Shire Boreen Point 19 0 0 1 0 1 0 0 0 0 1 0 1 0 0 0 1 5 Balance Sunshine Rural Verrierdale 16 0 0 1 0 1 0 1 0 0 0 0 0 0 1 0 1 5 Coast Region Balance Rural Noosa Shire Cootharaba 15 0 0 1 0 0 0 0 0 1 1 0 1 0 1 0 0 5 Balance Sunshine Rural Yandina Creek 14 0 0 0 0 0 0 1 1 0 1 0 0 0 1 0 1 5 Coast Region Balance Rural Noosa Shire Pomona 98 0 0 1 0 0 0 0 0 0 0 0 1 1 1 0 0 4 Balance Sunshine Other Beerwah 88 0 0 1 0 0 0 0 0 0 1 0 0 1 0 0 1 4 Coast Region Urban
Chapter 4: Results for Objective 2 133
Total Rural/ Total hot LGA name SSC name urban 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 cases spot category years Rural Noosa Shire Cooran 56 0 1 1 0 0 0 0 0 0 0 0 1 0 1 0 0 4 Balance Sunshine Mooloolah Other 41 0 0 1 0 0 0 1 0 0 0 0 0 1 0 0 1 4 Coast Region Valley Urban Sunshine Peregian Major 40 0 0 1 0 0 0 0 0 1 0 0 0 0 1 0 1 4 Coast Region Beach Urban Sunshine Rural Conondale 26 0 1 0 1 0 0 0 1 0 0 0 1 0 0 0 0 4 Coast Region Balance Lake Rural Noosa Shire 19 0 0 1 0 0 0 0 0 0 0 0 1 1 1 0 0 4 Macdonald Balance Sunshine Rural North Arm 11 0 0 1 0 0 0 1 0 0 0 0 0 0 1 0 1 4 Coast Region Balance Noosa North Rural Noosa Shire 9 0 0 1 0 1 0 0 0 0 1 0 0 0 1 0 0 4 Shore Balance Major Noosa Shire Noosa Heads 78 0 0 1 0 1 0 0 0 0 0 0 0 0 1 0 0 3 Urban Sunshine Glass House Other 76 0 0 0 0 0 0 0 1 0 0 0 0 1 0 0 1 3 Coast Region Mountains Urban Sunshine Other Landsborough 68 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 3 Coast Region Urban Moreton Bay Other Elimbah 39 1 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 3 Region Urban Moreton Bay Other Dayboro 34 0 0 1 0 0 0 0 0 0 0 0 0 1 0 1 0 3 Region Urban Sunshine Maroochy Other 21 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 1 3 Coast Region River Urban Rural Noosa Shire Kin Kin 19 0 0 1 0 0 0 0 0 0 0 0 1 0 1 0 0 3 Balance
Chapter 4: Results for Objective 2 134
Total Rural/ Total hot LGA name SSC name urban 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 cases spot category years Sunshine Rural Ninderry 14 0 0 0 0 0 0 1 1 0 0 0 0 0 1 0 0 3 Coast Region Balance Sunshine Rural Beerburrum 10 1 0 0 0 0 0 0 1 0 0 0 0 1 0 0 0 3 Coast Region Balance Rural Ipswich City Willowbank 9 0 0 0 0 0 0 0 1 1 0 0 0 0 0 1 0 3 Balance Sunshine Rural Valdora 6 0 0 0 0 0 0 0 1 0 0 0 0 0 1 0 1 3 Coast Region Balance Rural Noosa Shire Pinbarren 5 0 0 1 0 0 0 0 0 0 0 0 1 0 1 0 0 3 Balance Sunshine Other Coolum Beach 111 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 2 Coast Region Urban Gold Coast Major Ormeau 90 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 2 City Urban Sunshine Other Yandina 72 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 2 Coast Region Urban Gold Coast Other Jacobs Well 41 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 2 City Urban Sunshine Mount Other 31 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 2 Coast Region Coolum Urban Sunshine Peregian Major 20 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 2 Coast Region Springs Urban Rural Scenic Rim Harrisville 15 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 2 Balance Rural Noosa Shire Tinbeerwah 15 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 0 2 Balance Rural Ipswich City Purga 13 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 2 Balance
Chapter 4: Results for Objective 2 135
Total Rural/ Total hot LGA name SSC name urban 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 cases spot category years Moreton Bay Rural Donnybrook 12 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 2 Region Balance Black Rural Noosa Shire 10 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 2 Mountain Balance Rural Ipswich City Walloon 7 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 2 Balance Sunshine Mount Rural 6 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 2 Coast Region Mellum Balance Sunshine Major Weyba Downs 5 0 0 0 0 0 0 0 0 1 0 0 0 0 1 0 0 2 Coast Region Urban Total 1885 4 5 24 1 11 1 7 10 7 10 1 11 12 29 4 15
Chapter 4: Results for Objective 2 136
Chapter 5: Results for Objective 3
Chapter 5 includes the following manuscript:
Gyawali N, Murphy AK, Hugo LE, Devine GJ. A micro‐PRNT for the detection of Ross River virus antibodies in mosquito blood meals: a useful tool for inferring transmission pathways.
In this chapter, a laboratory technique was developed to aid field investigations of potential transmission pathways for RRV, by studying mosquito feeding patterns. This work was undertaken in concert with the field work undertaken in the following chapter (6). This technique scaled‐down and extended upon existing micro‐PRNT methodology to allow both the host origin, and host antibody status to RRV, to be determined from a single mosquito blood meal. This increases the information possible to be gained about mosquito‐host blood feeding relationships than can be gained using traditional methods.
This chapter describes this new approach and its application in examining the feeding behaviours of important mosquito vector species, along with assessing RRV exposure in vertebrate host species. The technique employed represents a novel, indirect approach to screening wildlife blood samples which can simultaneously obtain information on a) wildlife species fed on by specific vectors, and b) which wildlife species have evidence of previous RRV exposure.
Given the wide range of potential host species that could be involved in RRV transmission, this approach might assist in narrowing down the most important species to target in future studies. Particularly in the absence of broad wildlife blood screening studies, and/or sufficient experimental wildlife and vector infection studies, this approach represents a useful step forward toward clarifying the key vectors and hosts incriminated in RRV transmission, especially if applied in high risk environments, such as those identified in Chapter 4.
Chapter 5: Results for Objective 3 137
An additional paper produced as part of this objective, but not included in this thesis can be found in Appendix A: Interpreting mosquito feeding patterns in Australia through an ecological lens: an analysis of blood meal studies. This paper provides some additional context for Chapter 5, with a review existing blood meal studies performed in Australia to date, an assessment of their contribution to understanding vector‐borne disease risks, and identification of research gaps and future directions.
Chapter 5: Results for Objective 3 138 QUT Verified Signature
5.1. A micro‐PRNT for the detection of Ross River virus antibodies in mosquito blood meals: a useful tool for inferring transmission pathways
Narayan Gyawali1*, Amanda K. Murphy1, Leon E. Hugo1, Gregor J. Devine1
1 Mosquito Control Laboratory, QIMR Berghofer Medical Research Institute, Brisbane, Australia
* Correspondence:
Narayan Gyawali [email protected]
Running title: micro‐PRNT, a xenodiagnostic tool for arbovirus surveillance
Keywords: mosquito, antibody, seroprevalence, epidemiology, zoonosis, arbovirus
Chapter 5: Results for Objective 3 140
Abstract Introduction Many arboviruses of public health significance are maintained in zoonotic cycles with complex transmission pathways. The presence of serum antibody against arboviruses in vertebrates provides evidence of their historical exposure but reveals nothing about the vector‐reservoir relationship. Moreover, collecting blood or tissue samples from vertebrate hosts is ethically and logistically challenging. We developed a novel approach for screening the immune status of vertebrates against Ross River virus that allows us to implicate the vectors that form the transmission pathways for this commonly notified Australian arboviral disease. Methods A micro‐plaque reduction neutralisation test (micro‐PRNT) was developed and validated on koala (Phascolarctos cinereus) sera against a standard PRNT. The ability of the micro‐PRNT to detect RRV antibodies in mosquito blood meals was then tested using two mosquito models. Laboratory‐reared Aedes aegypti were fed, via a membrane, on sheep blood supplemented with RRV seropositive and seronegative human sera. Aedes notoscriptus were fed on RRV seropositive and seronegative human volunteers. Blood‐fed mosquitoes were harvested at various time points after feeding and their blood meals analysed for the presence of RRV neutralising antibodies using the micro‐PRNT. Results There was significant agreement of the plaque neutralization resulting from the micro‐PRNT and standard PRNT techniques (R2=0.65; P<0.0001) when applied to RRV antibody detection in koala sera. Sensitivity and specificity of the micro‐PRNT assay were 88% and 96%, respectively, in comparison with the standard PRNT. Blood meals from mosquitoes fed on sheep blood supplemented with RRV antibodies, and on blood from RRV seropositive humans neutralised the virus by ≥50% until 48 hr post feeding. The vertebrate origin of the blood meal was also ascertained for the same samples, in parallel, using established molecular techniques. Conclusions The small volumes of blood present in mosquito abdomens can be used to identify RRV antibodies and therefore host exposure to arbovirus infection. In
Chapter 5: Results for Objective 3 141
tandem with the accurate identification of the mosquito, and diagnostics for the host origin of the blood meal, this technique has tremendous potential for exploring RRV transmission pathways. It can be adapted for similar studies on other mosquito borne zoonoses.
Introduction Arthropod borne viruses (arboviruses) present a significant risk to public health globally. In recent decades, rapid urbanization and population growth have assisted the expansion of several viruses from having localised, rural, transmission cycles to being worldwide and urban problems [1]. Epidemiological cycles of many arboviruses, such as Ross River (RRV) and West Nile (WNV) incorporate complex transmission networks involving multiple vertebrate hosts and many vectors. Humans are not necessarily key components of these transmission networks, but increasing human travel, trade and deforestation bring humans into contact with sylvatic/enzootic cycles. This can stimulate arbovirus emergence, re‐emergence and spillover into human populations [2‐4]. A comprehensive knowledge of the transmission pathways of arboviruses is needed to effectively manage and respond to their emergence. Surveillance systems are needed to identify which mosquito species are responsible for transmission and which animals are acting as amplifying or reservoir hosts. However, the identification of amplifying hosts and transmission pathways remains extremely challenging. More than 75 arboviruses have been identified in Australia and a small number are associated with human infection [5]. Of these, RRV [6], Barmah Forest virus [7], WNV strain Kunjin [8], and the potentially fatal Murray Valley encephalitis virus [9] are of the greatest public health concern. RRV is the most commonly notified arboviral disease but multiple vectors and many potential vertebrate hosts make this a complex zoonosis. There is little empirical evidence regarding its key transmission cycles or encourage their spillover to the human population [10, 11]. One means of identifying likely vertebrate disease reservoirs is to demonstrate their historical exposure to disease by searching for virus‐specific antibodies in animal sera or tissues. Development of antibody is the major immune response to infection with parasites and pathogens including the arboviruses [12, 13].
Chapter 5: Results for Objective 3 142
While such serological evidence of infection does not prove that an animal is an amplifying host or key reservoir, it does allow the generation of hypotheses about probable pathways and is especially useful when combined with information on mosquito species and their host preference. Serological surveys of blood meals are likely to be more fruitful than the direct identification of viruses because vertebrates are only viraemic for a few days, only a small proportion of mosquitoes are virus positive and there is a diminishingly small probability that a captured mosquito will be carrying a virus positive mosquito blood meal. A tremendous sampling effort is therefore required to incriminate reservoir and vector pathways by virus isolation alone. The potential for screening mosquito blood meals for antibodies to dengue, Japanese encephalitis [14], and WNV [15] has been investigated previously but existing studies required the use of host‐specific conjugated antibodies. This is of little utility for the investigation of complex zoonoses like RRV where the hosts are myriad or unknown. The “gold standard” of serological tests is the Plaque Reduction Neutralisation Test (PRNT) [16]. It does not need prior knowledge of host origin but typically requires large quantities of sera or tissue; substantially larger than a typical mosquito blood meal (estimated to be 3 µl [17, 18]). We developed a micro‐PRNT [19, 20] to suit small sample volumes. In this alternative approach, we exploit the fact that vertebrate antibodies persist within mosquito blood meals for some time after the mosquito has fed on a seropositive host. We demonstrate that a micro‐PRNT technique can identify vertebrate RRV antibodies in small volumes of sera and mosquito blood meals. This has utility as part of an integrated xenodiagnostic approach that exploits the capture of single blood‐ fed mosquitoes to infer mosquito species, host preference and host exposure to disease. This will help prioritise potential transmission pathways for further study.
Methods Cells and virus
Vero cells (WHO vaccine strain) and the RRV strain T‐48 [20] were obtained from the WHO Collaborating Centre for Arbovirus Reference and Research at the
Chapter 5: Results for Objective 3 143
Queensland University of Technology. At the QIMR Berghofer Medical Research
o Institute (QIMRB), virus was propagated in Vero cells maintained in 5% CO2 at 37 C in RPMI‐1640 growth media (Sigma‐Aldrich, Missouri, USA), supplemented with L‐ glutamine (0.3 g/L), sodium bicarbonate (2 g/L), 10% (v/v) heat‐inactivated foetal calf serum (Invitrogen, USA) and 1% (v/v) PSG [(Penicillin (10,000 units)/Streptomycin (10 mg/mL)/ L‐glutamine (200 mM)); Sigma‐Aldrich, USA]. Virus stocks were frozen at ‐ 80°C.
Koala sera
Forty‐two koala sera, obtained from Endeavour Vets, Queensland, Australia (http://www.endeavourvet.com.au) were used to validate the micro‐PRNT. These samples were collected between 2015 to 2017 and stored at ‐80oC.
Mosquitoes
We used two insect models to validate the micro‐PRNT. An Ae. aegypti colony originated from Cairns, Australia in 2015, and was reared as previously described [21]. Adult mosquitoes were provided with 10% sugar solution ad libitum and an opportunity to feed on defibrinated sheep blood once per week. An Ae. notoscriptus colony was established from eggs collected in Brisbane, Australia during 2015 and maintained as above.
Development of the micro‐PRNT
All koala serum samples were tested for neutralising RRV antibodies using a conventional PRNT approach. Equal volumes of sera (200 µl), diluted 1:20 in serum‐ free RPMI‐1640, were mixed with an equal volume of 50 plaque forming unit RRV (1:800 stock RRV in RPMI‐1640) per well of a 12‐well tissue culture plate (Nunclon, Thermo Scientific, Australia). The virus‐sera mixtures were incubated at 37°C for 45 min and added to infect Vero cell monolayers and incubated for a further two hours to enable non‐neutralised virus to adsorb to cells. Following incubation, the virus‐ sera mixture was removed and 2 mL of 0.75% w/v carboxymethyl cellulose (CMC,
Sigma‐Aldrich) in RPMI 1640 was added. Plates were incubated at 37°C in 5% v/v CO2 for an additional 40 hr. The CMC/RPMI medium was then removed, and the cell
Chapter 5: Results for Objective 3 144
monolayers were fixed and stained with 0.05% w/v crystal violet (Sigma‐Aldrich) in formaldehyde (1% v/v) and methanol (1% v/v). The cell monolayers were then rinsed in tap water, and the plates inverted on a paper towel until dry. Plaques (clear zones in a purple cell monolayer) were counted. Reductions of total virus plaque numbers per well of ≥50% were considered to denote seropositive status [13].
All koala samples were also tested by a micro‐PRNT technique using just three µl of koala sera; the approximate volume of a mosquito blood meal [17]. Sera were diluted 1:20 in serum‐free RPMI‐1640. Equal volumes of sera (50 µl) were then mixed with equal volumes of 30 pfu RRV (1:160 stock RRV in RPMI‐1640) and added to duplicate wells (50 µl /well) of 96–well tissue culture plates (Nunclon, Thermo Scientific, Australia) containing a Vero cell monolayer. A virus density of 30 pfu per well in 96‐well plates allowed sufficient visual discrimination of plaques [23]. A volume of 200μl CMC/RPMI was added to each well following infection of the cell monolayer. Plates were scanned at 600 dpi resolution (HP Scanjet, Palo Alto, USA) and images were magnified before counting plaques manually.
The agreement between the percent plaque neutralisation from both the conventional and micro‐PRNT was determined by paired sample t‐test (n=42, each serum sample tested once with each PRNT protocol).
Preparation of blood fed mosquitoes
Blood cells from defibrinated sheep blood (Serum Australis, Manilla, NSW, Australia) were pelleted, washed three times in PBS and then reconstituted in either RRV seropositive or RRV seronegative human sera at a physiological proportion of 1: 0.82 blood cell: plasma. The blood was previously confirmed to be seronegative for RRV by conventional PRNT. Female Ae. aegypti mosquitoes (aged 3–5 days) were starved for 5 hr to increase their avidity and then offered seropositive or seronegative blood for 30 min via a membrane feeding apparatus [24]. Fully engorged and unfed mosquitoes from each treatment group were maintained separately at 27±1°C and 80% relative humidity and provided with 10% sucrose solution ad libitum. At 6, 12, 24, 36, 48, 60 and 72 hr post exposure to the infected and uninfected blood meals, fed and unfed mosquitoes (the latter used as controls) were harvested and stored at ‐80 °C.
Chapter 5: Results for Objective 3 145
In a separate experiment, to validate the micro‐PRNT on blood meals from mosquitoes that had fed directly on vertebrate hosts, Ae. notoscriptus were fed on the exposed arms of RRV seropositive and seronegative human volunteers (previously confirmed by conventional PRNT) for a period of 15 min. Fully engorged mosquitoes were selected for analysis, maintained and harvested as above. Remnant blood could be observed in mosquito abdomens until 60 hours (Fig. 5.1).
Informed, written consent was given for collection of volunteer blood samples and the direct feeding of mosquitoes on sero‐negative and sero‐positive humans (QIMR Berghofer Human Research Ethics approval P2273).
Fig. 5.1. Aedes notoscriptus at different stages of blood meal digestion.
Validation of the micro‐PRNT using blood‐fed mosquitoes
The blood meal volume obtained from a single field‐collected, blood‐fed mosquito is sufficient for the micro‐PRNT assay, but in order to ascertain the assay’s robustness against a range of host antibody titres and post‐feeding times, and facilitate the dilutions that these experiments required, we used larger volumes of mosquito‐derived blood in our validations. These were obtained by pooling three blood meals from engorged mosquitoes that had fed on the same antibody‐positive source. One pool was used for each post‐feeding time point tested.
Abdominal contents were diluted by 1:20, 1:40, 1:80 and 1:160 using serum‐ free RPMI 1640 supplemented with 1% PSG and 0.4% amphotericin B (Sigma‐Aldrich, USA). The final reported blood volume was calculated based on dilution of a 3 μl blood meal [17, 18]. A range of controls were also processed to assess the impacts of sero‐negative mosquito homogenates on the inhibition of plaque forming units. This entailed a comparison of 1) RRV alone 2) RRV plus unfed mosquito abdomens, and 3) RRV plus the abdomens of mosquitoes that contained sheep blood supplemented with seronegative human sera. In every case, the RRV pfu was kept constant.
To validate the ability of the micro‐PRNT to detect vertebrate anti‐RRV antibodies from single, blood fed mosquitoes, we tested Ae. notoscriptus that had
Chapter 5: Results for Objective 3 146
been fed on human seropositive or seronegative volunteers. Blood‐fed mosquitoes were harvested at various times post‐feeding and processed as above (abdominal contents expelled into serum‐free RPMI 1640 and adjusted to obtain a 20‐fold dilution). Unfed mosquitoes were included as a control. Each 96‐well plate was scanned and plaques were counted manually as described above.
Identification of host DNA
Sixteen blood meal samples harvested after feeding on RRV seropositive sheep blood or human blood (n = 8 for each) were used to demonstrate that host identification could be performed in parallel with the micro‐PRNT on the same blood samples. PCR amplification of Cytochrome b was performed as previously described [25].
Results Comparison of micro‐PRNT and PRNT
RRV plaques on cell layers stained 40 hr post‐incubation were clearly distinguishable and easily counted when plates were scanned and images enlarged. There was a significant correlation (R2=0.65; p<0.0001) between percent neutralization of RRV pfu noted in koala samples characterized by micro‐PRNT or PRNT techniques (Fig. 5.2). In comparison to the standard PRNT, the sensitivity and specificity of the micro‐PRNT was 88.2% and 96% respectively. Those differences in percent neutralisation determined between methods were not significant (p>0.05; paired samples t‐test).
Fig. 5.2. Plaque neutralisation demonstrated by standard PRNT or micro‐PRNT using koala serum samples.
Chapter 5: Results for Objective 3 147
Impacts of mosquito homogenates on the number of plaque forming units
Vero cells were inoculated with RRV alone, RRV mixed with homogenates of unfed mosquito abdomens and homogenates of abdomens containing RRV seronegative sheep blood. There was a small (10%) but significant decrease in RRV pfu when Vero cells were inoculated with RRV mixed with the latter two samples (Fig. 5.3).
Fig. 5.3. The effect of mosquito homogenates on plaque formation.
Validation of the assay using mosquito blood meals
Mosquito blood meals obtained from mosquitoes membrane‐fed on sheep blood mixed with anti‐RRV human antibodies neutralised RRV by ≥ 50% until 60 hr post blood feeding (Fig. 5.4). In contrast, there was no neutralization of blood meals obtained from mosquitoes fed with sheep blood supplemented with RRV seronegative human serum. These assays demonstrated that the micro‐PRNT is robust across a range of dilutions that are likely to represent varying antibody levels in the host.
Fig. 5.4. Percent neutralisation of RRV by vertebrate antibodies in mosquito blood meals harvested at different time points post blood‐feeding.
Guided by the results detailed in Fig. 5.4, we used the 1:20 dilution for all subsequent micro‐PRNTs on single blood meals from live hosts. Single Ae. notoscriptus blood meals obtained by feeding mosquitoes on a seropositive human volunteer were harvested at different time points. These blood meal preparations neutralised RRV pfu by ≥50% until 48 hr post feeding. (Fig. 5.5A). The limited
Chapter 5: Results for Objective 3 148
neutralising effects of sero‐negative blood meals (Fig. 5.5B) and un‐fed mosquito abdomens are included for comparison (Fig. 5.5C).
Fig. 5.5. Impact of post‐feeding times on neutralisation of RRV.
Host identification of blood meal samples
Sequencing of cytochrome b amplicon‐DNA from all 16 blood meal samples correctly identified (>95% nucleotide identity to cytochrome b sequences) the origin of the blood meal source (i.e. sheep from those experiments that had used membrane feeds, and human where mosquitoes had fed on volunteers).
Discussion Mosquito blood meals are a potentially useful resource for assessing antibody seroprevalence in vertebrates and inferring the RRV transmission pathways between vectors, disease reservoirs and humans. This study demonstrates that a micro‐PRNT using 96‐well plates has considerable utility for characterizing the antibody content of small blood volumes and is equivalent in sensitivity and specificity to the “gold standard” conventional PRNT method. In conjunction with the species identification of the blood‐fed mosquito, and the use of existing molecular tools to identify the host origin of the blood meal, this new diagnostic has the capacity to increase our understanding of the key pathways for the transmission of complex zoonotic arboviruses.
The accuracy of the micro‐PRNT for testing RRV antibodies in this study compared favourably with a conventional PRNT method when applied to our koala samples. Similarly, confluent results were reported for a micro‐PRNT tested against yellow fever virus in artificially spiked serum samples [19]. Although alternatives to the PRNT such as a VecTest‐inhibition assay and a biotin microsphere immunoassay have been used to identify pathogen‐specific antibodies in mosquito blood meals
Chapter 5: Results for Objective 3 149
[15]; both require host‐specific antibodies and the latter demands considerable laboratory resources. Our micro‐PRNT demonstrates sufficient sensitivity, specificity and utility for the determination of RRV antibodies in blood‐fed mosquitoes that have fed on any vertebrate.
Our micro‐PRNT could detect ≥50% RRV neutralisation by mosquito blood meals up to 36–48 hr post blood‐feeding. Although this is the first study to have identified RRV antibodies in mosquito blood meals, more general studies have shown that antibodies can survive in those environments. Hatfield (1988) identified Bovine Serum Albumin (BSA) specific antibodies using antibody‐captured ELISA from the hemolymph of Ae. aegypti up to 48 hr after feeding [26]. Irby and Apperson (1989) used an immunoblot technique to demonstrate that serum proteins from rodents and humans persisted in Ae. aegypti blood meals for 36 to 48 hr post feeding [27]. Anti‐BSA antibodies were detected 9 days after blood feeding in Anopheles stephensi [28] and human specific IgM and IgG were present in the blood meals of Ae. albopictus for 7 days [29]. These extended periods are surprising given that one would expect blood meals and their proteins to have been fully digested by then but the persistence of antibodies in mosquito blood meals will differ with species, ambient temperature, body size, initial concentration of antibodies, blood meal volume and the length of the gonotrophic cycle. In our study, the period of detectability (48–60 hr) corresponded to the period that blood was externally visible in the abdomen (Fig 5.1).
The small volumes of homogenate left from a single mosquito after execution of the micro‐PRNT allows for a parallel PCR amplification for identification of the blood meal source (the host). The literature commonly reports that the origin of host blood meals can be identified from as a little as 0.02 μl of blood [30]. In our proof of principle, 10 μl aliquots of 1:20 diluted homogenate recovered from the micro‐PRNT 60 hr post blood‐feeding were successfully amplified and sequenced to identify our experimental donors: sheep and humans.
Given the challenges involved in obtaining, trapping and screening wild animals for serum sampling and seroprevalence studies [13], the collection of blood fed mosquitoes is a useful means of sampling, with mosquitoes acting as an indirect
Chapter 5: Results for Objective 3 150
sampling tool or “flying syringe” for sampling inaccessible or ethically challenging hosts. In terms of pathway incrimination, a single mosquito will yield information on vectors, host preference and the disease exposure of that host [11]. That information, especially when combined with risk modelling [31] can be used to identify those transmission pathways of greatest importance within the host community. Various studies have observed the potential for insect blood meals to detect pathogen specific antibodies [26, 32–34], but none have developed high throughput methods suitable for application to transmission ecologies involving unknown reservoirs.
Human health can be affected by infectious diseases of wildlife living close to human habitation. The risks are increasingly common in Australia and elsewhere because of increasing encroachment of the human population on diverse mosquito habitats and the adaptation of pathogen reservoir species to urbanized environments [35, 36]. Dengue, Hantavirus, Lyme disease, Zika, avian influenza, and rabies are examples of globally endemic zoonoses that have emerged from human encroachment into rural or sylvatic habitats [37, 38]. Similarly, RRV is a major public health risk in Australia, maintained in a diverse range of hosts and vectors and undergoing an expansion in range to the Pacific Islands [39].
Infectious diseases are also a concern for wildlife conservation, particularly those already threatened by habitat loss and exploitation. Surveillance of wild animals for infection or disease commonly involves trapping or killing animals for direct sampling of blood and tissues. This can be difficult, expensive, dangerous and sometimes unethical. The technique demonstrated here is not only applicable to RRV reservoir identification but also to other arboviruses and infectious agents which have complex transmission cycles and a range of vertebrate hosts.
There was ~10% inhibition of virus pfu by mosquito tissue homogenates (Fig. 5.3). However, this inhibition was minimal compared to the threshold used to determine positive neutralization (>50% reduction). One possible explanation of this observation is that some component of abdominal tissue may have an inhibitory effect on virus replication.
As for all serum or tissue collection techniques, reliance on blood‐fed mosquitoes as a sampling tool will be subject to sampling bias. Mosquitoes may be
Chapter 5: Results for Objective 3 151
differentially attracted to diseased hosts [40] or to species that are uncharacteristically abundant at any single point in time. Different mosquito trapping techniques have differential vector specific targets, so using one particular trap type might miss key vector species. In terms of serology, there may be considerable cross reaction between some virus antibodies [12]. In this case RRV is likely to cross‐react with the closely related alphavirus, Barmah Forest virus, whose epidemiology in Queensland remains significant [2]. Finally, those reservoirs inferred by blood meal analysis may not be the key amplifying host, and the mosquitoes incriminated may not be the key vector. For example, dengue antibodies are commonly found in Culex spp. mosquitoes during epidemics, but those mosquitoes do not transmit the disease. Nonetheless their blood meals may still identify the host and its seroprevalence rate [14]. For all of these reasons, the various components of the pathways implicated by the micro‐PRNT technique and attendant host identifications must be interpreted and prioritised in the light of all the available knowledge on the ecology of the disease.
Conclusions
The value of the micro‐PRNT lies in its ability to detect anti‐virus antibodies from mosquito blood meals obtained from any vertebrate host. When coupled with molecular identification of the host by DNA amplification and sequencing, valuable information on vector‐host relationships, wildlife seroprevalence rates and zoonotic transmission cycles can be inferred [14]. This novel xenodiagnostic offers an alternative “flying syringe” approach for serum sampling and for monitoring seroprevalence in animals. The use of this assay to characterise the blood meals of mosquitoes collected from the field in Brisbane is now underway.
Acknowledgements
This work was supported by funds from the Mosquito Control Laboratory, QIMR Berghofer and by the Mosquito and Arbovirus Research Committee (MARC) which is an independent Australian organization funded by local government, government agencies, industry and scientific institutions. We thank Prof. John Aaskov and Dr. Francesca Frentiu (Queensland University of Technology) for their advice and for the supply of Vero cells and the RRV isolate.
Chapter 5: Results for Objective 3 152
Authors’ Contributions
AM, LEH, NG and GD conceived the project. NG, AM and LEH carried out the laboratory experiments. NG drafted the paper. GD, LEH, AM reviewed and revised the manuscript. All authors contributed to preparation of the final version and agreed to its submission.
Authors’ Disclosure Statement
The authors declare that they have no competing financial interests.
Chapter 5: Results for Objective 3 153
References 1. Gould E, Pettersson J, Higgs S, Charrel R, de Lamballerie X. Emerging arboviruses: Why today?. One Health. 2017;4:1–13. doi:10.1016/j.onehlt.2017.06.001. 2. Gyawali N, Bradbury RS, Aaskov JG, Taylor‐Robinson AW. Neglected Australian arboviruses: quam gravis?. Microbes Infect. 2017;19(7‐8):388–401. doi:10.1016/j.micinf.2017.05.002. 3. Rückert C, Ebel GD. How Do Virus‐Mosquito Interactions Lead to Viral Emergence?. Trends Parasitol. 2018;34(4):310–321. doi:10.1016/j.pt.2017.12.004. 4. Mayer SV, Tesh RB, Vasilakis N. The emergence of arthropod‐borne viral diseases: A global prospective on dengue, chikungunya and zika fevers. Acta Trop. 2017;166:155–163. doi:10.1016/j.actatropica.2016.11.020. 5. Centers for Disease Control and Prevention (CDC). CDC Arbovirus Catalogue. Retrieved October 01 2019, from Centers for Disease Control and Prevention (CDC), https://wwwn.cdc.gov/Arbocat/Default.aspx 6. Fraser JR. Epidemic polyarthritis and Ross River virus disease. Clin Rheum Dis. 1986;12(2):369–388. 7. Phillips DA, Murray JR, Aaskov JG, Wiemers MA. Clinical and subclinical Barmah Forest virus infection in Queensland. Med J Aust. 1990;152(9):463–466. 8. Muller D, McDonald M, Stallman N, King J. Kunjin virus encephalomyelitis. Med J Aust. 1986;144(1):41–42. 9. French EL. Murray Valley encephalitis isolation and characterization of the aetiological agent. Med J Aust. 1952;1(4):100–103. 10. Claflin SB, Webb CE. Ross River Virus: Many Vectors and Unusual Hosts Make for an Unpredictable Pathogen. PLoS Pathog. 2015;11(9):e1005070. doi:10.1371/journal.ppat.1005070. 11. Stephenson EB, Peel AJ, Reid SA, Jansen CC, McCallum H. The non‐human reservoirs of Ross River virus: a systematic review of the evidence. Parasit Vectors. 2018;11(1):188. doi:10.1186/s13071‐018‐2733‐8. 12. Gyawali N, Taylor‐Robinson AW, Bradbury RS, Pederick W, Faddy HM, Aaskov JG. Neglected Australian Arboviruses Associated With Undifferentiated Febrile Illnesses. Front Microbiol. 2019;10:2818. doi:10.3389/fmicb.2019.02818. 13. Gyawali N, Taylor‐Robinson AW, Bradbury RS, Potter A, Aaskov JG. Infection of Western Gray Kangaroos (Macropus fuliginosus) with Australian Arboviruses Associated with Human Infection. Vector Borne Zoonotic Dis. 2019;10.1089. doi:10.1089/vbz.2019.2467. 14. Barbazan P, Palabodeewat S, Nitatpattana N, Gonzalez JP. Detection of host virus‐reactive antibodies in blood meals of naturally engorged mosquitoes. Vector Borne Zoonotic Dis. 2009;9(1):103–108. doi:10.1089/vbz.2007.0242 15. Komar N, Panella NA, Young GR, Basile AJ. Methods for detection of West Nile virus antibodies in mosquito blood meals. J Am Mosq Control Assoc. 2015;31(1):1–6. doi:10.2987/14‐6468R.1 16. Kuno G. Serodiagnosis of flaviviral infections and vaccinations in humans. Adv Virus Res. 2003;61:3‐65. doi: 10.1016/s0065‐3527(03)61001‐8
Chapter 5: Results for Objective 3 154
17. Konishi E. Size of blood meals of Aedes albopictus and Culex tritaeniorhynchus (Diptera: Culicidae) feeding on an unrestrained dog infected with Dirofilaria immitis (Spirurida: Filariidae). J Med Entomol. 1989;26(6):535‐8. 18. Jeffery GM. Blood meal volume in Anopheles quadrimaculatus, A. albimanus and Aedes aegypti. Exp. Parasitol. 1956;5(4):371‐375. 19. Simões M, Camacho LA, Yamamura AM, Miranda EH, Cajaraville AC, da Silva Freire M. Evaluation of accuracy and reliability of the plaque reduction neutralization test (micro‐PRNT) in detection of yellow fever virus antibodies. Biologicals. 2012;40(6):399–404. doi:10.1016/j.biologicals.2012.09.005. 20. Lee SS, Phy K, Peden K, Sheng‐Fowler L. Development of a micro‐neutralization assay for ebolaviruses using a replication‐competent vesicular stomatitis hybrid virus and a quantitative PCR readout. Vaccine. 2017;35(41):5481–5486. doi:10.1016/j.vaccine.2017.03.019 21. Doherty RL, Carley J, Mackerras MJ, Marks EN. Studies of Arthropod‐borne Virus Infection in Queensland: III. Isolation and Characterisation of Virus Strains from Wild‐caught Mosquitoes in North Queensland. Aust J Exp Biol Med Sci. 1963;41(1):17‐39. 22. Hugo LE, Monkman J, Dave KA, Wockner LF, Birrell GW, Norris EL, et al. Proteomic biomarkers for ageing the mosquito Aedes aegypti to determine risk of pathogen transmission. PLoS One. 2013;8(3):e58656. doi:10.1371/journal.pone.0058656. 23. Borges MBJ, Kato SE, Damaso CR, Moussatché N, da Silva Freire M, Passos SRL, et al. Accuracy and repeatability of a micro plaque reduction neutralization test for vaccinia antibodies. Biologicals. 2008;36(2):105–110. doi:10.1016/j.biologicals.2007.07.001. 24. Hugo RLE, Stassen L, La J, Gosden E, Winterford C, Viennet E, et al. Vector competence of Australian Aedes aegypti and Aedes albopictus for an epidemic strain of Zika virus. PLoS Negl Trop Dis. 2019;13(4):e0007281. doi:10.1371/journal.pntd.0007281. 25. Gyawali N, Taylor‐Robinson AW, Bradbury RS, et al. Identification of the source of blood meals in mosquitoes collected from north‐eastern Australia. Parasit Vectors. 2019;12(1):198. doi:10.1186/s13071‐019‐3455‐2. 26. Hatfield PR. Detection and localization of antibody ingested with a mosquito bloodmeal. Med Vet Entomol. 1988;2(4):339–345. doi:10.1111/j.1365‐ 2915.1988.tb00206.x 27. Irby WS, Apperson CS. Immunoblot analysis of digestion of human and rodent blood by Aedes aegypti (Diptera: Culicidae). J Med Entomol. 1989;26(4):284– 293. doi:10.1093/jmedent/26.4.284 28. Lackie AM, Gavin S. Uptake and persistence of ingested antibody in the mosquito Anopheles stephensi. Med Vet Entomol. 1989;3(3):225–230. doi:10.1111/j.1365‐2915.1989.tb00220.x. 29. Tesh RB, Chen WR, Catuccio D. Survival of albumin, IgG, IgM, and complement (C3) in human blood after ingestion by Aedes albopictus and Phlebotomus papatasi. Am J Trop Med Hyg. 1988;39(1):127–130. doi:10.4269/ajtmh.1988.39.127.
Chapter 5: Results for Objective 3 155
30. Service M, Voller A, Bidwell D. The enzyme‐linked immunosorbent assay (ELISA) test for the identification of blood‐meals of haematophagous insects. Bull Entomol Res. 1986;76(2):321‐330. 31. Louis VR, Phalkey R, Horstick O, Ratanawong P, Wilder‐Smith A, Tozan Y, et al. Modeling tools for dengue risk mapping ‐ a systematic review. Int J Health Geogr. 2014;13:50. doi:10.1186/1476‐072X‐13‐50. 32. Vaughan JA, Azad AF. Passage of host immunoglobulin G from blood meal into hemolymph of selected mosquito species (Diptera: Culicidae). J Med Entomol. 1988;25(6):472–474. doi:10.1093/jmedent/25.6.472. 33. Minoura H, Chinzei Y, Kitamura S. Ornithodoros moubata: host immunoglobulin G in tick hemolymph. Exp Parasitol. 1985;60(3):355–363. doi:10.1016/0014‐ 4894(85)90042‐6. 34. Ben‐Yakir D, Fox CJ, Homer JT, Barker RW. Quantification of host immunoglobulin in the hemolymph of ticks. J Parasitol. 1987;73(3):669–671. 35. Australian Government. Our North, Our Future: White Paper on Developing Northern Australia; 2015. Available at: http://northernaustralia.gov.au/files/files/NAWP‐FullReport.pdf [accessed October 02, 2019]. 36. Gyawali N, Bradbury RS, Aaskov JG, Taylor‐Robinson AW. Neglected Australian Arboviruses and Undifferentiated Febrile Illness: Addressing Public Health Challenges Arising From the 'Developing Northern Australia' Government Policy. Front Microbiol. 2017;8:2150. doi:10.3389/fmicb.2017.02150. 37. Baylis M. Potential impact of climate change on emerging vector‐borne and other infections in the UK. Environ Health. 2017;16(Suppl 1):112. doi:10.1186/s12940‐017‐0326‐1. 38. Vorou RM, Papavassiliou VG, Tsiodras S. Emerging zoonoses and vector‐borne infections affecting humans in Europe. Epidemiol Infect. 2007;135(8):1231– 1247. doi:10.1017/S0950268807008527. 39. Lau C, Aubry M, Musso D, Teissier A, Paulous A, Despres P, et al. New evidence for endemic circulation of Ross River virus in the Pacific Islands and the potential for emergence. Int J Infect Dis. 2017;57:73–76. doi:10.1016/j.ijid.2017.01.041 40. De Moraes CM, Stanczyk NM, Betz HS, Pulido H, Sim DG, Read AF, et al. Malaria‐induced changes in host odors enhance mosquito attraction. Proc Natl Acad Sci USA. 2014;111(30):11079–11084. doi:10.1073/pnas.1405617111.
Chapter 5: Results for Objective 3 156
Tables and Figures
Chapter 5: Results for Objective 3 157
Figure 5.1. Aedes notoscriptus at different stages of blood meal digestion.
Immediately after feeding (0 hr), mosquitoes were fully engorged. After 72 hr, the blood meal could no longer be seen.
Chapter 5: Results for Objective 3 158
Figure 5.2. Plaque neutralisation demonstrated by standard PRNT or micro‐PRNT using koala serum samples.
Correlation of percent reduction in plaque forming units as measured by standard PRNT (x‐ axis) and micro‐PRNT (y‐axis) techniques.
Chapter 5: Results for Objective 3 159
Figure 5.3. The effect of mosquito homogenates on plaque formation.
Homogenates of mosquito abdomens reduce the number of plaque forming units by approximately 10%, when compared with Vero cells inoculated with RRV alone (*p<0.05, calculated by one‐way ANOVA test).
Chapter 5: Results for Objective 3 160
Figure 5.4. Percent neutralisation of RRV by vertebrate antibodies in mosquito blood meals harvested at different time points post blood‐feeding.
Blood meals obtained from RRV positive sera and harvested at different times post‐feeding were diluted at four different concentrations and tested in duplicate. Antibodies continued to neutralise ≥ 50% of plaques at all dilutions, until 60 hr post feeding.
Chapter 5: Results for Objective 3 161
Figure 5.5. Impact of post‐feeding times on neutralisation of RRV.
Blood meals from an RRV seropositive human volunteer (A), blood meals from an RRV seronegative human volunteer (B) and unfed mosquito abdomens as a control (C). Antibody positive blood meals continued to neutralise ≥ 50% of plaques until 48 hours post‐feeding.
Chapter 5: Results for Objective 3 162
Chapter 6: Results for Objective 4
This chapter includes work resulting from a 9‐month field study conducted in Brisbane, as a collaboration primarily between QIMRB’s Mosquito Control Laboratory and Griffith University’s Disease Ecology Group, with contributions and input also provided by Queensland Health and QUT. The results for this objective are outlined in the following manuscript:
Murphy AK, Graham M, Gyawali N, Skinner EB, Jansen CC, Shivas MA, Onn MB, Rabellino A, Hu W, Hafner LM, Frentiu FD, Devine GJ. Mosquitoes as flying syringes: investigation of Ross River virus epidemiology and host seroprevalence using mosquito blood meals.
This chapter describes the use of a combination of vertebrate and vector field surveys and laboratory analysis of mosquito blood meals to incriminate specific vectors and hosts in RRV transmission. Field studies took place in 4 study sites in Brisbane, and were timed to encompass the peak season for RRV transmission in SEQ. Investigations incorporated the use of the new xenodiagnostic (micro‐PRNT) approach that was described in Chapter 5. Although the collection of sufficient data to conclusively link specific vectors, hosts and RRV was outside the scope of this study, it provided an initial proof‐of‐concept to inform future investigations.
The findings of this chapter contribute to the understanding of RRV transmission dynamics in the endemic city of Brisbane, and provide a framework for expanded investigations of RRV ecology. Although this was a relatively small‐scale study in terms of number of sites and number of blood meal samples, I show that while both vector and host species distributions differ slightly between rural and urban areas of Brisbane, the feeding patterns of key vectors were uniform across sites. Additionally, using the micro‐PRNT technique (Chapter 5), I obtained information on the RRV serostatus of several bird species for the first time. This
Chapter 6: Results for Objective 4 163
demonstrates the importance of exploring the specific transmission factors that operate in different local habitat types, and the value of using multi‐disciplinary approaches to define and quantify the complex drivers of arboviral and zoonotic diseases.
An additional paper produced as part of this objective, but not included in this thesis can be found in Appendix A: Interpreting mosquito feeding patterns in Australia through an ecological lens: an analysis of blood meal studies. This paper reports the findings for additional, related objectives of the field study – outlined in the field survey protocol. It describes the different community assemblages of vector and host species across a range of Brisbane study sites, and explored how the balance of species in different sites might contribute to RRV disease risk.
In addition, a detailed protocol was created to guide the field study, and to clarify the roles of participating institutions. This can be found in Appendix B: RRV field survey protocol. This field study included a number of broad objectives, encompassing Objective 4 of this research: to investigate relationships between vector‐host interactions and RRV transmission in urban and suburban locations through the assessment of mosquito feeding behaviours, relative to available host species.
Chapter 6: Results for Objective 4 164 QUT Verified Signature
6.1. Mosquitoes as flying syringes: using mosquito blood meals to implicate vectors and hosts in Ross River virus transmission
Amanda K. Murphy1,2*, Melissa Graham1, Narayan Gyawali1, Eloise B. Skinner3, Cassie C. Jansen4, Martin A. Shivas5, Michael B. Onn5, Andrea Rabellino6, Wenbiao Hu7, Louise M. Hafner2, Francesca D. Frentiu2, Gregor J. Devine 1
1 Mosquito Control Laboratory, QIMR Berghofer Medical Research Institute, Brisbane, Australia
2 School of Biomedical Sciences, and Institute for Health and Biomedical Innovation, Queensland University of Technology, Brisbane, Australia
3 Environmental Futures Research Institute, Griffith University, Brisbane, Australia
4 Communicable Diseases Branch, Queensland Department of Health, Brisbane, Australia
5 Brisbane City Council, Field Services, Brisbane, Australia
6 Signal Transduction Laboratory, QIMR Berghofer Medical Research Institute, Brisbane, Australia
7 School of Public Health and Social Work, Queensland University of Technology, Brisbane, Australia
* Corresponding author, e‐mail: [email protected]
Keywords: Ross River virus, arbovirus, zoonosis, vectors, hosts, blood feeding
Chapter 6: Results for Objective 4 166
Abstract Ross River virus (RRV) is a zoonotic pathogen with a complex transmission cycle involving multiple vector and reservoir host species. Despite regular outbreaks causing substantial morbidity around Australia each year, the most important vectors and hosts which drive transmission cycles are not known. Marsupials have long been thought to play a key role as amplifying hosts, although their role in maintaining the virus in urban areas is unknown. We combined our wildlife and vector surveys with the use of both established and novel molecular laboratory assays to investigate transmission pathways across 4 urban and suburban sites in Brisbane, Australia. Over a 9‐month period encompassing the annual peak RRV season, we surveyed 111 wildlife species and collected 6,420 mosquitoes. Host origin of 204 mosquito blood meals were identified through amplification and sequencing of the cytochrome b gene. In addition, blood meals were simultaneously analysed for antibodies to RRV using a novel micro‐plaque reduction neutralisation assay. Across 10 different mosquito species collected from urban and suburban parks of Brisbane, humans were the dominant source of blood meal for all mosquito species (76%), followed by birds (18%). This contrasts with previous studies in Brisbane and other states in which humans represented a relatively small proportion of meals. Further, RRV seroprevalence was 20% in the 131 blood meal samples of human origin, and 16% in 37 samples from bird species. Seroprevalence studies, particularly in non‐human vertebrates, are challenging to conduct due to ethical and practical constraints, but this study provides a novel, non‐invasive method of estimating seroprevalence in both humans and wildlife. Additionally, we present the first investigation of RRV serostatus in several bird species. These findings contribute to the understanding of RRV transmission dynamics in Brisbane, and provide a framework for expanded investigations of RRV ecology.
Chapter 6: Results for Objective 4 167
Introduction Understanding the ecological factors which facilitate transmission of vector‐ borne zoonotic disease is essential to the design of disease prevention and control strategies. However, this can be a challenge for multi‐host pathogens, and particularly where the hosts responsible for pathogen maintenance (reservoir hosts) may differ from those which facilitate epidemic transmission (amplification hosts) [1, 2]. These relationships are especially important to monitor as environments, human behaviours and landscapes undergo change. Environmental change has provided several zoonotic pathogens with opportunities to invade new environments – and these dynamics have become particularly important to understand in the context of ongoing global urbanisation [3‐5]. Despite its necessity for successful disease management strategies, knowledge about these relationships is incredibly difficult to obtain.
Ross River virus (RRV) is a zoonotic arbovirus transmitted across a range of environments in Australia and the Pacific Islands, which causes regular outbreaks of polyarthritic disease across Australia [6]. The mosquito vector and reservoir host species mediating RRV’s continued circulation, and contributing to human outbreaks, are uncertain. The virus can infect several mosquito and vertebrate host species, with marsupials suspected to represent a major, but not sole, reservoir [7‐9]. Humans, other mammals and birds could also play a role in maintaining RRV transmission [7]. Further, at least 10 mosquito vector species can transmit RRV across both salt and freshwater habitats across different bioclimatic and geographic regions [10]. Historically considered a rural pathogen, RRV cases associated with metropolitan areas of Australian cities have gradually increased over the past 2 decades [10‐13]. This changing trend in RRV disease patterns is important to monitor for maintaining preparedness and management of future epidemics.
In sub‐tropical South East Queensland, where there are high rates of RRV disease, incidence rates differ between urban and rural areas [Murphy et al. Submitted manuscript, Chapter 4]. Hence, this region provides an appropriate setting to explore new tools for better understanding RRV transmission ecology. However, previous studies have focused on assessing only one of many one components of
Chapter 6: Results for Objective 4 168
transmission at a time (such as vector incrimination studies, or host sero‐surveys). Consequently, the complexity of interactions between these components – and how these relate to human disease risk – may not be captured. Studies which examine vector, host and viral interactions simultaneously have greater potential to elucidate the key interactions underlying pathogen transmission.
Our study used a multidisciplinary approach to investigate RRV vector‐host feeding relationships in the city of Brisbane, in south east Queensland, Australia. We conducted surveys of both mosquitoes and vertebrates across urban and suburban areas of Brisbane, and applied a unique combination of laboratory techniques to determine wildlife host origins and RRV seroprevalence of mosquito blood meals in these areas. We hypothesized that vectors and hosts implicated in RRV transmission might differ between urban and suburban areas of Brisbane. Our aim was to compare host feeding patterns of vectors between urban and suburban areas, and to explore how vector‐host dynamics might contribute to RRV epidemiology in this region.
Methods Study location
Queensland is Australia’s third most populated state, with 4.9 million inhabitants, including 1.1 million in the capital city of Brisbane. Brisbane is a coastal city, with a sub‐tropical climate. Average minimum and maximum monthly temperatures are 12‐23◦C in winter, and 20‐29◦C in the summer months between November and January [14]. A seasonal increase in rainfall occurs from December‐ March, which precedes the annual peak in RRV notifications, typically between February and May [14, 15]. However, earlier occurrence of rainfall during some years can be associated with an increase in cases, typically from October onwards [16]. During the peak transmission season, the virus circulates widely across Brisbane city, although the spatial trend changes each year [12, Murphy et al. Submitted manuscript, Chapter 4]. Sites located within 4 suburbs of Brisbane (Fig. 6.1): Herston, Red Hill, Chermside West and Pullenvale were selected as study sites for characterisation of potential vector and host species. Two sites comprised inner city
Chapter 6: Results for Objective 4 169
parks (in Herston and Red Hill) and two comprised outer suburban forest reserves (in Chermside West and Pullenvale). Geographic visualisation of the study sites employed the State Suburb (SSC) geographical boundary [17], and graphical maps were created using ArcMap 10.6 (ESRI, Redlands, CA, USA).
Monthly surveys of vertebrate and mosquito species were conducted in each suburb between September 2017 and May 2018, in conjunction with a wider study surveying 5 additional Brisbane suburbs. The results of the wider survey are under review elsewhere [Skinner et al. Appendix C]. The 4 suburbs chosen for this study were public recreation areas selected based on their feasibility for blood fed mosquito collection, namely those that provided adequate harbourage for resting blood fed females, either as natural or artificial mosquito resting sites, including vegetation, bridges or drains. Each site was also required to contain suitable habitat for reservoir host species (particularly wallabies and possums), and evidence of mosquito breeding (such as larvae observed in water bodies), and be of sufficient size to encompass 3 x 100 m x 100 m survey transects per site (marked using hand‐held GPS devices; Garmin‐62S), with at least 150 m separating each transect. Surveys of mosquitoes and vertebrate hosts were conducted in each transect, with data from each transect aggregated for sites, on each survey date.
Figure 6.1. Map of Australia and Brisbane.
Mosquito & vertebrate surveys
Mosquitoes were collected over a 9‐month period: September 2017 to May 2018, which overlapped with the vertebrate surveys. Various methods were used to characterise mosquito species, and maximise the likelihood of catching blood fed mosquitoes (S6.2 Table), including light traps (Pacific Biologics, Scarborough,
Australia) baited with CO2 as dry ice pellets, Biogent’s Gravid Aedes Traps (Pacific Biologics, Australia), aspiration using the ProkoPack device (original model, from John W. Hock Company, U.S.A.), and larval collection by dip method. Light traps were set
Chapter 6: Results for Objective 4 170
in each site once per month before dusk and collected the following morning. A single light trap was located within one of three transects in each study site, and rotated between transects each month. Traps were set 1.5 m off the ground adjacent to a water source. GAT traps were set in different locations across each site each month, usually with two GATs per transect. Aspiration was conducted once/month for each site by 1 collector for 5 minutes/transect. To maximise efficient use of time and resources, aspiration was focused around artificial resting sites such as underneath bridges that traversed water, inside large drains and culverts, and surrounding vegetation. Larval surveys were conducted sporadically across sites, only during months when mosquito breeding was observed. Adult mosquitoes captured were sorted on a cold table, and identified to species level using morphological keys. Blood‐ fed mosquitoes were separated and stored at ‐80◦C for additional processing, described below under Laboratory Analysis Techniques.
Vertebrate surveys were undertaken in conjunction with mosquito surveys (October 2017‐March 2018) to estimate feeding indices and) involved monthly dawn and dusk surveys in each site as described in Skinner et al. [Appendix C]. Briefly, surveys were non‐invasive and recorded the number of individuals and species of all vertebrates observed within a specified time (30 minutes for dawn surveys, 30 person minutes for dusk surveys). Vertebrate surveys took place in the same transects as for the mosquito sampling.
Blood meal analyses
Abdomens of single blood fed mosquitoes were removed under a dissection microscope, and homogenised into 70 µl of tissue culture media (Roswell Park Memorial Institute‐1640) supplemented with 1% PSG [(Penicillin (10,000 units)/Streptomycin (10 mg/mL)/ L‐glutamine (200 mM)); Sigma‐Aldrich, USA] and 0.4% amphotericin B (Sigma‐Aldrich, USA). Each sample was centrifuged at 10,000 × g for 10 min and the supernatant removed. A 55 µl volume of supernatant was transferred to a sterile Eppendorf tube to be tested for the presence of RRV neutralising antibodies using a micro‐PRNT assay, while the remaining (~10µl) volume was retained in a separate tube for molecular analysis to identify the blood meal host
Chapter 6: Results for Objective 4 171
origin. All samples were frozen at ‐80◦C before use in either DNA extraction or micro‐ PRNT assays.
Amplification and sequencing of a 457 base‐pair region of the cytochrome b (Cytb) mitochondrial DNA gene extracted from blood meal homogenates was performed according to methodology described by Flies et al. [13]. Briefly, vertebrate and mammalian primers for Cytb were selected and validated before use on blood or serum samples from known host species including: humans, sheep and koalas. DNA was extracted from each 10 µl blood meal sample using Purelink genomic DNA extraction kit, purified and amplified using Polymerase Chain Reaction (PCR) with a touchdown cycling sequence to minimize non‐specific amplification. Purified DNA was then sequenced in‐house using QIMRB’s DNA sequencing facilities, and the sequences matched to those in the NCBI nucleotide database.
The presence of neutralising RRV antibodies in the mosquito blood meal was tested by a scaled‐down version of a Plaque Reduction Neutralisation Test (PRNT), or micro‐PRNT as described by Gyawali et al. [Chapter 5]. Briefly, wells of a 96‐well tissue culture plate (Nunclon, Thermo Scientific, Australia) were inoculated with Vero cells to produce a confluent cell monolayer. The following day, each blood meal homogenate was combined with 30 pfu of RRV (T48 genotype (Doherty et al. 1963), provided by the WHO Collaborating Centre for Arbovirus Reference and Research, Queensland University of Technology), and the total volume split and added to wells in duplicate, along with the addition of positive and negative controls. After incubation for 2 hours at 37◦C, the virus‐sera mixture was removed and 2 mL of 0.75% w/v carboxymethyl cellulose (CMC, Sigma‐Aldrich)/RPMI 1640 was added, followed
◦ by further 40‐hour incubation at 37 C in an atmosphere of 5% v/v CO2/air. Wells were then emptied of liquid and stained with 0.05% w/v crystal violet (Sigma‐Aldrich) in formaldehyde (1% v/v) and methanol (1% v/v).
After 24 hours of staining, plates were rinsed in cold water and allowed to dry. Using magnified photographs of each well, the visible number of plaques/well were counted by two independent and blinded operators. The number of plaques counted by each were combined and the average of the two taken as the final result. Blood
Chapter 6: Results for Objective 4 172
meals which inhibited the number of viral plaques by ≥ 50% compared to the control (RRV virus alone) were considered RRV reactive/seropositive.
Data analysis
Feeding ratios for the number of non‐human blood meals obtained in relation to the abundance of blood meal host species were calculated as:
Observed number of blood meals per vertebrate species Feeding ratio Expected number of blood meals per vertebrate species where, the expected number of blood meals = the total number of blood meals obtained multiplied by the relative proportion (abundance) of each species. The resulting ratio will be equal to 1 if the observed proportion of blood meals from a given species is equivalent to its relative abundance in the wildlife population surveyed. A value greater or lesser than 1 would indicate either increased or decreased proportion of blood meals were obtained from this species relative to its overall abundance.
Results Mosquito species captured
A total of 6,420 adult mosquitoes from 21 species were captured across Brisbane field sites (Table 6.1). Mosquito abundance was greatest between November and April, and the majority of blood fed species were caught between January and April (S6.1 Fig.). The number and diversity of species differed between the urban parks and the suburban forests, with greater overall abundance and diversity of species in the forest reserves (17 species) compared to the city parks (14 species), as well as differences in the proportions of species present (Fig. 6.2). The dominant species captured in suburban forests were Aedes procax (55%), Culex annulirostris (24%) and Ae. vigilax (14%) while in urban parks, the highest proportions were Cx. sitiens (32%), Cx. annulirostris (26%) and Cx. quinquefasciatus (11%). The most abundant mosquito species across all sites was Ae. procax (46% of all species), followed by Cx. annulirostris (25%) and Ae. vigilax (13%).
Chapter 6: Results for Objective 4 173
Ten species of blood fed mosquitoes totalling 240 blood fed mosquitoes were captured, primarily through aspiration (Tables 6.1, S6.2). The majority of blood fed mosquitoes were captured from urban sites (n=217, 90%) compared to the suburban sites (n=23, 10%). This greater proportion from urban sites was influenced by one urban site (Herston) which was highly productive of blood fed mosquitoes. Despite this, blood fed mosquitoes were generally representative of abundant species across Brisbane, though included greater relative proportions of Cx. sitiens (51%) and Cx. quinquefasciatus (19%), and comparatively lower proportions of Ae. procax (6%), Ae. vigilax (2%) and Cx. annulirostris (15%) (Table 6.1, Fig. 6.2).
Table 6.1. Adult mosquitoes captured across Brisbane.
Figure 6.2. Adult mosquitoes captured in urban and suburban sites.
Vertebrate species recorded
A total of 111 non‐human vertebrate species were identified during vertebrate surveys across the four sites, which were summarised into 12 broad species groups (Fig. 6.3). The majority of species diversity was represented by birds (92), followed by placental mammals (11) and marsupials (8). However, in terms of total abundance, the dominant number and proportion of species overall were placental mammals (52%), followed by birds (45%), and marsupials (3%). The Herston site had the highest vertebrate abundance overall, due to the presence of a large flying fox colony (of ≥ 500 individuals/survey) resident at that site. This increased its the proportion of placental mammals to 77%, compared with birds (22%), which were dominant at all other sites. The other notable difference between sites was the distribution of marsupials, with wallabies observed only in the suburban sites only, while possums were present in every site but were more abundant in the urban parks. Overall, the
Chapter 6: Results for Objective 4 174
urban parks had higher abundance of all species groups, and a greater proportion of placental mammals (60%, of which 57% were flying foxes and 3% dogs) than suburban forests (25%, of which 21% were flying foxes and 3% dogs). Urban parks also had a lower proportion of bird species (37%) compared to suburban forests (72%). Marsupials comprised ~3% of total species in both urban and suburban sites.
Figure 6.3. Relative abundance of vertebrate species across Brisbane sites.
Blood meal origins of mosquitoes
Of the 240 blood meals investigated in laboratory analyses, the host origin was successfully identified from 204 + 1 that appeared to comprise a mixed blood meal (human and bird origin) (Tables 6.2, S6.3). Identified blood meal hosts comprised 19 vertebrate host species, 13 of which were birds. Humans were the dominant origin of mosquito blood meals across all sites (155/204 blood meals;76%), followed by birds (39/204; 19%). After birds, 5/204 (2.5%) blood meals were from non‐human placental mammals (3 domestic cats and 2 black flying foxes), and 4/204 (1%) from marsupials (3 brushtail possums, and 1 red‐necked wallaby). The human dominance as blood meal host origin was consistent across each site and each mosquito species, with all but one species identified as having at least one bloodmeal from humans. Bird feeding was identified in five mosquito species, including Ae. vigilax, (1/4 blood meals identified), Cx. annulirostris (9/37), Cx. australicus (2/2), Cx. quinquefasciatus (13/45) and Cx. sitiens (14/122) (Table 6.2; Fig. 6.4).
Comparison of non‐human blood meals across sites suggested a greater proportion of blood meals from birds, domestic cats and wallabies in suburban forests than in urban parks; however, given the low number of samples from suburban sites, this requires further verification (Fig. 6.4). Across all sites, around half of blood meals originating from birds (20 of 39) were from the noisy miner (Manorina melanocephala), followed by the Australian magpie (Gymnorhina tibicen) and the
Chapter 6: Results for Objective 4 175
Australasian figbird (Sphecotheres vieilloti), with 3 blood meals each from the latter two (Table S6.4). Of the bird blood meals, 34/39 originated from the Herston site, and the remaining 5 were from Chermside West and Pullenvale.
Table 6.2. Host origins of mosquito blood meals.
Figure 6.4. Host origins of mosquito blood meals by urban and suburban collection sites.
Blood meal origins versus vertebrate abundance
Low sample sizes of non‐human blood meals limited a robust assessment of relative feeding patterns. Because humans observed in study sites during vertebrate surveys were not included within vertebrate counts, their availability as blood meals in comparison to other vertebrate species is unknown, although their presence was informally observed across all sites during most surveys. While the sample size of blood meals of non‐human origin was relatively low (n=48), especially from suburban sites (n=8), we compared this total number of non‐human blood meals against the expected number, based on relative abundance of non‐human vertebrates from the surveys. This comparison indicated that the number of blood meals from all non‐ human vertebrates were lower than expected (feeding ratios of 0.4 for birds, 0.05 for placental mammals, and 0.6 for marsupials). Hence, the small number of marsupial blood meals captured (n=4) might be considered relatively high given their low abundance in the surveys. Similarly, the relatively high proportion of blood meals obtained from birds (n=39) was relatively low when compared with abundance.
RRV serostatus in blood meal hosts across Brisbane
Chapter 6: Results for Objective 4 176
Of 204 blood meal samples assessed by micro‐PRNT, 27 (13%) were lost/unsuccessful due to contamination during the cell culture process, leaving 177 samples that returned either an RRV‐reactive (seropositive) or RRV‐nonreactive (seronegative) result (Table 6.3). Of these, 36 (20%) were seropositive. These included 26 bloodmeals of human origin, 6 from birds, 2 from placental mammals (both domestic cats), and 2 from marsupials. From the 10 blood‐fed mosquito species analysed, 4 had fed on RRV seropositive hosts: Ae. procax, Cx. annulirostris, Cx. quinquefasciatus, and Cx. sitiens. In total, 31/36 seropositive samples (86%) were from the urban sites (30 from Herston, 1 from Red Hill), while 5/36 were from the suburban site, Pullenvale (Fig. 6.5, S6.5 Table). None of the 3 samples tested from Chermside West (all of bird origin) tested positive. Despite noisy miners being the most abundant bird species fed on by mosquitoes, none of the 19 blood meals tested from noisy miners were seropositive. Single seropositive samples were identified from an Australian raven (Corvus coronoides), a bush stone‐curlew (Burhinus grallarius), an eastern whipbird (Psophodes olivaceus), an Australian white Ibis (Threskiornis molucca), an Australian magpie (Gymnorhina tibicen), and a Tawny frogmouth (Podargus strigoides) (S6.5 Table).
Figure 6.5. RRV seropositive hosts in urban and suburban sites.
Table 6.3. Seropositive hosts identified from mosquito blood meals.
Chapter 6: Results for Objective 4 177
Discussion To understand and respond to the increasing RRV threat in Australia and internationally, improved characterisation of the determinants of human infection is required. This study examined RRV vector‐host relationships in Brisbane, and compared how these might differ between urban and suburban areas. We observed urban‐suburban differences in vector and host species composition, which could relate to differences in RRV transmission pathways. Across both urban and suburban sites, there was a striking pattern of human dominance in blood meal origins of all mosquito species, and these blood meals of human origin had an overall RRV seroprevalence of 20%. Bird species were the second most common group of vertebrates fed on by mosquitoes across Brisbane, and in these species the seroprevalence was 16%.
Urban suburban differences in vertebrate and mosquito species, and the association of urban spaces with altered abundance and diversity of species, has been documented in previous studies [19‐22]. The impact of these differences is not often explored or documented alongside epidemiological investigations. In this respect, our study offered a unique approach in its combination of ecological surveys and host‐feeding analyses. Our findings emphasise the value of obtaining data on relative host abundance to support understanding of vector feeding patterns. This has also been observed in other disease systems, such as for West Nile virus [1]. Our study is also the first to specifically compare vector‐host interactions between urban and suburban areas of Brisbane. The two previous studies in Brisbane identified non‐ human mammals and birds as dominant hosts [23, 24]. Our findings also suggest that some mosquito species whose feeding patterns have primarily been studied in rural environments may have different feeding patterns in urban areas; for example, Culex sitiens, which has previously been considered as primarily ornithophilic. This also fits with the observation of Jansen et al (2009), who found that feeding rates on birds differed for Cx. quinquefasciatus between different Australian cities (95% in Cairns and 50% in Sydney) [25].
This is also the first demonstration of use of a micro‐PRNT technique to indirectly assess RRV serostatus in wildlife hosts. Although an RRV seropositive or
Chapter 6: Results for Objective 4 178
negative blood meal does not definitively determine whether or not a particular vertebrate host is involved in RRV transmission, this information can contribute to hypothesises of which hosts have potential to play a role. The majority of current theories about RRV hosts are based on seroprevalence surveys, along with a few experimental infection studies comparing relative viraemia levels among small samples of wildlife [7].
Although a small sample, our seroprevalence data generated from micro‐PRNT assays presents the first investigation of RRV exposure for the 12 species tested. For marsupials, 2/4 tested were seropositive, which although a very small sample, is in line with other studies of marsupial seroprevalence (Skinner et al. Submitted manuscript). We also found 2/3 blood meals from domestic cats, and 0/2 from black flying foxes were seropositive. Further studies of the viraemic response and experimental infection studies are needed for birds, placental mammals and marsupials, along with a more detailed understanding of how seasonal vector‐host interactions in particular environments contribute to viral amplification and spread. However, in the absence of these, the micro‐PRNT is a useful alternative to aid host hypotheses, or to narrow down host targets for further investigation.
At all sites, humans were the dominant mammal identified from blood meals, despite other mammals being equally, if not more accessible to mosquitoes – especially in the suburban forests. This human dominance of blood meals is notable when compared to previous studies in Brisbane and across Australia, where humans comprised a low proportion of blood meals (11.6% by Kay, 2007; 17% by Flies 2016) which may reflect differences in trapping location and methods, and shows how much this influences the results. For example, Kay et al. placed mosquito traps primarily in domestic back yards and found that dogs dominated mosquito blood meal origins, possibly due to the proximity of these traps to domestic pets [23]. It may also reflect the larger biomass of humans influencing mosquito attraction, perhaps especially with their use of outdoor urban spaces for commuting, exercise or leisure activities. It is notable that 6/7 mosquito species that are competent to transmit RRV showed a feeding pattern that included both humans and at least one other vertebrate species. Further, three of the four vector species which fed on
Chapter 6: Results for Objective 4 179
seropositive humans also fed on other seropositive hosts. These included Ae. procax, Cx. quinquefasciatus and Cx. sitiens. This may provide some evidence for vectors that might be important in virus spillover/spillback events. Notwithstanding the labour‐ intensive collection of blood fed female mosquitoes, analysis of an expanded number of blood meal samples has substantial potential to further reveal potential links between specific vectors, humans and other vertebrates in supporting RRV transmission. Ideally, future studies would also consider additional factors that could also influence blood feeding choices of mosquitoes, such as vector genetics, host availability & biomass [26].
The relatively high proportion of blood meals originating from birds is comparable to that of previous studies in urban Brisbane (18% by Kay 2007, 32% by Jansen 2009), and lower than a recently study in urban Adelaide (49% of city‐caught blood meals by Flies 2016), although the sample size of the latter study was low, and this result may have been influenced by their trap sites being close to swamps [23, 24, 27]. In our vertebrate surveys, birds comprised 45% of all vertebrates counted, indicating that the proportion of blood meals originating from birds is lower than would be expected based on abundance alone. This could also reflect the influence of host biomass of mosquito feeding choices, and humans being a more attractive host source, but a potential preference of feeding on humans versus birds might also vary with different mosquito species. The defensive behaviours of birds (and other species) against mosquito biting may have also influenced greater human feeding. Notably, abundant bird species such as Rainbow lorikeets (Trichoglossus moluccanus) and Torresian crows (Corvus orru) did not appear to be fed on regularly, but further samples would be needed to confirm this.
Another factor potentially influencing feeding on specific birds could be their level of activity around the times of day that mosquitoes are most active. Lorikeets in particular are often actively engaged in feeding or transiting during dawn and dusk, compared with Noisy Miners who were often observed sleeping in trees during our evening vertebrate surveys. Alongside the influence of bird behaviour and activity levels on mosquito biting rates, host competence data for specific bird species would be valuable to estimate the possible effect of mosquito feeding on either
Chapter 6: Results for Objective 4 180
amplification or dilution of the virus. This would require experimental infection studies to test, and to date this has only been performed for a single bird species, the Little Corella, which was demonstrated to transmit RRV to Cx. annulirostris [28].
There was a surprisingly low rate of feeding on non‐human placental mammals (2.5%) compared to what has been observed in previous blood feeding studies (57% in Kay 2007, 9% in Jansen 2009, 66% in Flies 2016) [23, 24, 27]. Our results comprised a total of three blood meals originating from domestic cats. Previous studies in Brisbane have identified domestic cats as a relatively low proportion of blood meals (2% in Kay 2007, and 1% in Jansen 2009) [23, 24]. Neither cats nor dogs are thought to be effective amplifiers of RRV, and it’s possible they have a dilution effect as hosts, but this requires confirmation [29]. Our finding of proportionally low feeding on non‐ human placental mammals is likely due to trapping mosquitoes in public recreation areas, where, with the exception of flying foxes, these animals had relatively low abundance. This again illustrates the effect that trapping location can have on blood feeding results – and highlights the importance of using broad sampling methodology within a given environment to enable accurate characterisation of host feeding patterns. Given that the majority (90%) of our blood meals came from the urban site, Herston, which included a large flying fox colony (where the species Pteropus alecto, P. scapulatus and P. poliocephalus roost together), we were surprised to find the relatively low proportion of blood meals (1%) originating from flying foxes. However, it was noteworthy that the two blood meals obtained were both from the black flying fox, and not from the grey‐headed flying fox which is similarly abundant in this site. A 1997 Brisbane study adjacent to another flying fox colony with mixed species present also found blood meals to originate only from black flying foxes (16/20 or 80% blood meals tested from Ve. funerea) [30], while a 2007 study in the same site found 6/21 or 29% of blood meals from Cx. annulirostris, Cx. sitiens and Ae. vigilax originated from flying foxes, but did not specify which species [31].
Black flying foxes could be of particular interest to explore in further studies, as their ecological characteristics as least circumstantially highlight them as a species of interest. This species has been implicated as carriers of other zoonotic viruses, such as Hendra virus, have become abundant in urban areas over the last 20 years, travel
Chapter 6: Results for Objective 4 181
widely in search of food which may bring them into proximity with vector mosquitoes, and have a seasonal breeding cycle which may align with the seasonal peak in RRV notifications [7, 32]. Although the two black flying fox blood meals we tested were found to be RRV seronegative, this is a small sample, and additional sampling from this species would be valuable to improve evidence of its status as a host.
Both urban and suburban sites had a surprisingly low proportion of marsupial blood meals (2%). Yet the feeding index showed that the few samples obtained were proportional to overall marsupial abundance (3%). Still, it was disappointing not to be able to investigate additional samples of marsupials given their proposed role in RRV transmission. The low proportion of feeding observed for marsupials may be simply due to chance or an artefact of our collection methods, but if additional evidence were to confirm this as a consistent trend, it could potentially challenge the assumption that RRV is maintained in urban areas by marsupials. This is not the first study to cast doubt on their role in transmission (Hill et al 2009). However, this contrasts with evidence from seroprevalence studies of possums which suggest that possums appear to be regularly infected with RRV. Additional sampling would be useful in future to further address the role of marsupials in urban and suburban areas of Brisbane – and the framework presented here may allow an efficient approach to both investigate feeding patterns and RRV exposure for marsupials, as well as facilitate comparisons with a range of other species not yet investigated [7].
The main limitation of this study was the inability of its sample size to make statistical inferences on vector‐host relationships, particularly between urban and suburban blood meals. Nevertheless, we present some preliminary results that show promise for an expanded study. Our approach of using blood fed mosquitoes to indirectly sample wildlife presents a viable option to assess exposure in a range of wildlife species which may not be practical to conduct specific trapping for, and helps to narrow into which species are most important to focus on in future studies. We also show that abundance alone is not an accurate indicator of mosquito feeding patterns or transmission pathways. More detailed studies of vector, host and virus interactions are critical to progress our understanding of RRV transmission pathways,
Chapter 6: Results for Objective 4 182
and to inform more targeted prevention strategies. However, our study provides a proof of concept for a novel approach that could yield valuable insights into RRV ecology.
Overall, these investigations of potential drivers of RRV disease contribute to the understanding of RRV transmission dynamics in Brisbane, and provide a framework for expanded investigations of RRV ecology. The collection of data on multiple aspects of RRV ecology, including vector‐host species composition, vector‐ host feeding patterns, and host seroprevalence, provides an efficient means to better understand the RRV transmission cycle. This is critical to the development of prevention and control strategies for RRV, but may also have applications to other zoonotic diseases where the key vectors and hosts involved in transmission are not known.
Acknowledgements
This work was supported by funds from the Mosquito Control Laboratory, QIMR Berghofer and by the Mosquito and Arbovirus Research Committee (MARC) which is an independent Australian organization funded by local government, government agencies, industry and scientific institutions. We thank Dr Jonathan Darbro of Metro North Public Health Unit; Mr Keith Rickart of the Queensland Department of Health, Communicable Diseases Branch; Prof. John Aaskov, formerly of QUT; and Mosquito Control Lab members: Dr Leon Hugo, Dr Oselyne Ong, Dr Brian Johnson, Dr Gordana Rasic and Mr Igor Filipovic for their helpful input and advice.
Chapter 6: Results for Objective 4 183
Authors’ Contributions
AM, ES, GD and FF conceived the project. AM, ES, MO and AR carried out the field work, with input from CJ, MS and GD. AM, MG, NG and AR completed the laboratory analyses. AM drafted the paper. ES, CJ, MS, MO, FF, WH and LH aided interpretation. All authors reviewed and revised the final manuscript and agreed to its submission.
Chapter 6: Results for Objective 4 184
References 1. Kilpatrick, A.M., et al., Host heterogeneity dominates West Nile virus transmission. Proc Biol Sci, 2006. 273(1599): p. 2327‐33. 2. Kilpatrick, A.M. and S.E. Randolph, Drivers, dynamics, and control of emerging vector‐borne zoonotic diseases. Lancet, 2012. 380(9857): p. 1946‐55. 3. Chaves, L.S.M., et al., Abundance of impacted forest patches less than 5 km(2) is a key driver of the incidence of malaria in Amazonian Brazil. Sci Rep, 2018. 8(1): p. 7077. 4. Fornace, K.M., et al., Association between Landscape Factors and Spatial Patterns of Plasmodium knowlesi Infections in Sabah, Malaysia. Emerg Infect Dis, 2016. 22(2): p. 201‐8. 5. Kilpatrick, A.M., Globalization, land use, and the invasion of West Nile virus. Science, 2011. 334(6054): p. 323‐7. 6. Claflin, S.B. and C.E. Webb, Ross River Virus: Many Vectors and Unusual Hosts Make for an Unpredictable Pathogen. PLoS Pathog, 2015. 11(9): p. e1005070. 7. Stephenson, E.B., et al., The non‐human reservoirs of Ross River virus: a systematic review of the evidence. Parasit Vectors, 2018. 11(1): p. 188. 8. Harley, D., A. Sleigh, and S. Ritchie, Ross River virus transmission, infection, and disease: a cross‐disciplinary review. Clin Microbiol Rev, 2001. 14(4): p. 909‐32, table of contents. 9. Flies, E.J., et al., Another Emerging Mosquito‐Borne Disease? Endemic Ross River Virus Transmission in the Absence of Marsupial Reservoirs. BioScience, 2018. 68(4): p. 288‐293. 10. Russell, R.C., Ross River virus: ecology and distribution. Annu Rev Entomol, 2002. 47: p. 1‐31. 11. Brokenshire, T., et al., A cluster of locally‐acquired Ross River virus infection in outer western Sydney. N S W Public Health Bull, 2000. 11(7): p. 132‐134. 12. Jansen, C.C., et al., Epidemiologic, Entomologic, and Virologic Factors of the 2014–15 Ross River Virus Outbreak, Queensland, Australia. Emerg Infect Dis, 2019. 25(12). 13. Lindsay, M., et al., An outbreak of Ross River virus disease in Southwestern Australia. Emerg Infect Dis, 1996. 2(2): p. 117‐20. 14. Australian Bureau of Meteorology Climate Data Online. Australian Bureau of Meteorology. [Accessed November 28, 2019]; Available from: http://www.bom.gov.au/jsp/ncc/climate_averages/temperature/index.jsp. 15. Notifiable Conditions Reports. Queensland Department of Health. [Accessed September 23, 2018]; Available from: https://www.health.qld.gov.au/clinical‐ practice/guidelines‐procedures/diseases‐ infection/surveillance/reports/notifiable/annual/default.asp. 16. National Notifiable Diseases Surveillance System. Australian Government. [Accessed September 23, 2018]; Available from: http://www9.health.gov.au/cda/source/cda‐index.cfm. 17. Australian Statistical Geography Standard (ASGS): Volume 3 ‐ Non ABS Structures. 2019. 18. Wilson, D.B. and A.S. Msangi, An estimate of the reliability of dipping for mosquito larvae. East Afr Med J, 1955. 32(2): p. 37‐9.
Chapter 6: Results for Objective 4 185
19. Gottdenker, N.L., et al., Anthropogenic land use change and infectious diseases: a review of the evidence. Ecohealth, 2014. 11(4): p. 619‐32. 20. Arunachalam, N., et al., Eco‐bio‐social determinants of dengue vector breeding: a multicountry study in urban and periurban Asia. Bull World Health Organ, 2010. 88(3): p. 173‐84. 21. Li, Y., et al., Urbanization increases Aedes albopictus larval habitats and accelerates mosquito development and survivorship. PLoS Negl Trop Dis, 2014. 8(11): p. e3301. 22. Vannavong, N., et al., Effects of socio‐demographic characteristics and household water management on Aedes aegypti production in suburban and rural villages in Laos and Thailand. Parasit Vectors, 2017. 10(1): p. 170. 23. Kay, B.H., et al., Mosquito feeding patterns and natural infection of vertebrates with Ross River and Barmah Forest viruses in Brisbane, Australia. Am J Trop Med Hyg, 2007. 76(3): p. 417‐23. 24. Jansen, C.C., et al., Blood sources of mosquitoes collected from urban and peri‐ urban environments in eastern Australia with species‐specific molecular analysis of avian blood meals. Am J Trop Med Hyg, 2009. 81(5): p. 849‐57. 25. Jansen, C.C., et al., Arboviruses isolated from mosquitoes collected from urban and peri‐urban areas of eastern Australia. J Am Mosq Control Assoc, 2009. 25(3): p. 272‐8. 26. Stephenson, E.B., et al., Interpreting mosquito feeding patterns in Australia through an ecological lens: an analysis of blood meal studies. Parasit Vectors, 2019. 12(1): p. 156. 27. Flies, E.J., et al., Regional Comparison of Mosquito Bloodmeals in South Australia: Implications for Ross River Virus Ecology. J Med Entomol, 2016. 53(4): p. 902‐910. 28. Kay, B.H., et al., Experimental infection of vertebrates with Murray Valley Encephalitis and Ross River viruses. Arbovirus Res. Aust., 1986 (Proceedings Fourth Symposium, May 6‐9, 1986): p. 71‐75. 29. Boyd, A.M. and B.H. Kay, Assessment of the potential of dogs and cats as urban reservoirs of Ross River and Barmah Forest viruses. Aust Vet J, 2002. 80(1‐2): p. 83‐6. 30. Ryan, P.A., et al., Investigation of gray‐headed flying foxes (Pteropus poliocephalus) (Megachiroptera: Pteropodidae) and mosquitoes in the ecology of Ross River virus in Australia. Am J Trop Med Hyg, 1997. 57(4): p. 476‐82. 31. van den Hurk, A.F., I.L. Smith, and G.A. Smith, Development and evaluation of real‐time polymerase chain reaction assays to identify mosquito (Diptera: Culicidae) bloodmeals originating from native Australian mammals. J Med Entomol, 2007. 44(1): p. 85‐92. 32. Peel, A.J., et al., Synchronous shedding of multiple bat paramyxoviruses coincides with peak periods of Hendra virus spillover. Emerg Microbes Infect, 2019. 8(1): p. 1314‐1323.
Chapter 6: Results for Objective 4 186
Tables and Figures
Chapter 6: Results for Objective 4 187
Figure 6.1. Map of Australia and Brisbane. a) Map of Australia, showing Brisbane’s location; b) Map of Brisbane showing study sites in urban (in blue: Red Hill (R) and Herston (H)) and suburban (in green: Pullenvale (P) and Chermside West (C)) locations. Brisbane’s northern and southern suburbs are divided by the Brisbane River. The black point indicates Brisbane’s city centre.
Chapter 6: Results for Objective 4 188
Table 6.1. Adult mosquitoes captured across Brisbane.
Urban parks Surburban forest Number blood fed Mosquito species Red Chermside Total Proportion Herston Pullenvale (% of all Hill West BF) Ae. aculeatus 8 8 0.1% Ae. notoscriptus* 27 79 39 34 179 2.8% 9 (4%) Ae. procax* 39 82 2318 542 2981 46.4% 14 (6%) Ae. theobaldi 5 5 0.1% Ae. vigilax* 41 50 720 13 824 12.8% 4 (2%) Ae. vittiger 1 8 39 3 51 0.8% An. annulipes* 3 3 9 3 18 0.3% 2 (1%) An. bancroftii 2 2 0.0% Cq. linealis* 4 2 6 0.1% Cq. xanthogaster 4 4 0.1% Cx. annulirostris* 184 135 1165 108 1592 24.8% 37 (15%) Cx. australicus* 5 4 9 0.1% 2 (1%) Cx. orbostiensis 4 2 15 45 66 1.0% Cx. pullus 7 1 8 0.1% 4 (2%) Cx. quinquefasciatus* 105 27 1 8 141 2.2% 45 (19%) 122 Cx. sitiens* 341 51 82 4 478 7.4% (51%) L. halifaxii 6 6 0.1% 1 (0.4%) Ma. uniformis* 4 4 0.1% Mi. elegans 2 2 0.0% Ve. funerea* 2 7 9 0.1% Ve. Marks sp.52 27 27 0.4% 240 Total 1212 5208 6420 100% (100%) * Species demonstrated as competent RRV vectors in laboratory studies.
Chapter 6: Results for Objective 4 189
Figure 6.2. Adult mosquitoes captured in urban and suburban sites. a) Total proportions of mosquitoes captured across urban and suburban sites; and b) proportions of blood fed mosquito species captured across urban and suburban sites.
Chapter 6: Results for Objective 4 190
Figure 6.3. Relative abundance of vertebrate species across Brisbane sites. a) Average number of individuals of each species group per survey month across the 4 study sites; and b) Relative proportions of each species group across site types.
Chapter 6: Results for Objective 4 191
Table 6.2. Host origins of mosquito blood meals.
Hosts species origin of blood meals
Bird Placental mammal Marsupial Unclear Bird Black Domestic Brushtail Red necked Bird and/or Mosquito species Human Unknown Total spp.^ flying fox cat possum wallaby Human Ae. notoscriptus* 8 1 9 Ae. procax* 6 2 1 5 14 Ae. vigilax* 1 1 2 4 An. annulipes* 2 2 Cx. annulirostris* 9 21 1 6 37 Cx. australicus 2 2 Cx. pullus 3 1 4 Cx. quinquefasciatus* 13 27 1 1 3 45 Cx. sitiens* 14 86 1 2 19 122 L. halifaxii 1 1 Total 39 155 2 3 3 1 1 36 240 Proportion 19.1% 76.0% 1.0% 1.5% 1.5% 0.5% 0.5% 100% * Species demonstrated to be competent RRV vectors ^ See S6.4 Table for specific bird species identified
Chapter 6: Results for Objective 4 192
Figure 6.4. Host origins of mosquito blood meals by urban and suburban collection sites. a) Number of blood meals by study site, and b) proportion of blood meals by site type.
Chapter 6: Results for Objective 4 193
Figure 6.5. RRV seropositive hosts in urban and suburban sites.
The number of RRV seropositive blood meals originating from each host species is shown, and whether captured in an urban or suburban study site.
Chapter 6: Results for Objective 4 194
Table 6.3. Seropositive hosts identified from mosquito blood meals. RRV seropositive host species Red‐ Blood‐fed mosquito Black Domestic Brushtail Bird Human necked Total species flying fox cat possum wallaby Ae. notoscriptus* 0 Ae. procax* 1 1 1 3 Ae. vigilax* 0 An. annulipes* 0 Cx. annulirostris* 2 2 Cx. australicus 0 Cx. pullus 0 Cx. quinquefasciatus* 1 4 1 6 Cx. sitiens* 5 19 1 25 L. halifaxii 0 Total seropositive 6 26 0 2 1 1 36 samples Total seronegative 31 105 2 1 2 0 141 samples No result 2^ 24^ 0 0 0 0 26^ Total samples tested 37 131 2 3 3 1 177 Proportion seropositive 16% 20% 0% 40% 33% 100% 20% * Species demonstrated to be competent RRV vectors ^ Plus one additional sample that was from both bird and human origin = 27 samples in total for which no result was obtained
Chapter 6: Results for Objective 4 195
Supporting Information
S6.1 Figure. Mosquitoes captured across Brisbane by month.
S6.2 Table. Mosquitoes captured across Brisbane by each trap method.
S6.3 Table. Host species origins of blood fed mosquitoes.
S6.4 Table. Bird species origins of blood fed mosquitoes.
S6.5 Table. RRV seroprevalence of mosquito blood meal hosts across urban and suburban sites.
Chapter 6: Results for Objective 4 196
S6.1 Figure. Mosquitoes captured across Brisbane by month.
Chapter 6: Results for Objective 4 197
S6.2 Table. Mosquitoes captured across Brisbane by each trap method.
Urban sites Suburban sites Total blood‐ Total Chermside fed Herston Red Hill Pullenvale mosquitoes Trap method West mosquitoes Mechanical aspiration 463 3 26 20 512 210 BG 2 2 10 GAT 8 88 6 5 107 0 Light Trap 299 334 4399 746 5778 20 Reared larvae 17 3 1 21 0 Total mosquitoes 770 442 4434 774 6420 captured Total blood‐fed 205 12 4 19 240 mosquitoes
Chapter 6: Results for Objective 4 198
S6.3 Table. Host species origins of blood fed mosquitoes.
Species group Species name Common name Genbank accession numbers* Burhinus grallarius Bush stone‐curlew DQ385219.1 Corvus coronoides Australian raven AF197837.1 Cracticus nigrogularis Pied butcherbird MF077450.1 Grallina cyanoleuca Magpie lark KP036677.1 Gymnorhina tibicen Magpie EF173688.1 Manorina flavigula Yellow‐throated miner AY488354.1 Bird Manorina melanocephala Noisy miner AF197859.1 Phylidonyris albifrons Honeyeater spp. FJ499020 Podargus strigoides Tawny frogmouth JQ353838.1 Psophodes olivaceus Eastern Whipbird FJ821139.1 Sphecotheres vieilloti Fig Bird JN614878.1 Threskiornis molucca Australian white ibis FJ498970 Felis catus Domestic cat AB194817.1 Homo sapiens Human MN540529.1 Placental mammal Pteropus alecto Black flying fox KJ532442.1 Macropus rufogriseus Red necked wallaby EF368027.1 Trichosurus vulpecula Brushtail Possum AF152857.1 Marsupial Trichosurus vulpecula Brushtail Possum DQ385219.1
* Based on identification of the mitochondrial Cytb gene
Chapter 6: Results for Objective 4 199
S6.4 Table. Bird species origins of blood fed mosquitoes.
Bird Bird species species total
Bush Willy wagtail Australian Eastern Honeyeater Magpie Noisy Pied Tawny stone‐ Fig Bird Ibis Magpie or raven Whipbird sp. lark miner* butcherbird frogmouth Mosquito species curlew butcherbird^ Ae. notoscriptus Ae. procax Ae. vigilax 1 1 An. annulipes Cx. annulirostris 1 1 6 1 9 Cx. australicus 1 1 2 Cx. pullus Cx. quinquefasciatus 1 2 1 1 8 13 Cx. sitiens 2 1 1 2 2 4 1 1 14 L. halifaxii Grand Total 2 1 1 3 1 3 3 1 20 1 2 1 39 * Unable to distinguish from the closely related Yellow‐throated miner ^ Unable to distinguish these related species due to limited sequences available in NCBI’s database
Chapter 6: Results for Objective 4 200
S6.5 Table. RRV seroprevalence of mosquito blood meal hosts across urban and suburban sites.
Birds Placental mammals Marsupials
Willy Bush Honey Pied Black Red Study site Australian Eastern Fig Magpie Noisy Tawny wagtail or Domestic Brushtail stone‐ eater Ibis Magpie butcher Human flying necked raven Whipbird bird lark miner* frogmouth butcher cat Possum Total type curlew spp. bird fox wallaby bird^ Number of blood meals tested Urban 2 1 0 3 1 2 2 1 16 1 2 1 124 2 1 3 0 162 Suburban 0 0 1 0 0 1 0 0 3 0 0 0 7 0 2 0 1 15 Total tested 2 1 1 3 1 3 2 1 19 1 2 1 131 2 3 3 1 177 Number of blood meals RRV seropositive Urban 1 1 0 0 0 1 1 0 0 0 1 0 24 0 1 1 0 31 Suburban 0 0 1 0 0 0 0 0 0 0 0 0 2 0 1 0 1 5 Total RRV 1 1 1 0 0 1 1 0 0 0 1 0 26 0 2 1 1 36 seropositive * Unable to distinguish from the closely related Yellow‐throated miner ^ Unable to distinguish these related species due to limited sequences available in NCBI’s database
Chapter 6: Results for Objective 4 201
Chapter 7: General discussion
Arboviral diseases are a global public health problem, with multifaceted drivers. Many have limited prevention or treatment strategies available, and their burdens are expected to rise in the coming decades. Opportunities for emergence and spread of arboviruses will likely continue to occur through the processes of globalisation, urban development, human travel and trade, and environmental modification. Despite growing recognition of the role of environmental change on arboviral disease risk, understanding of the drivers of transmission across different ecological settings, and strategies for their management, are lacking.
The research presented in this thesis sought to improve understanding of two medically important arboviruses, and the underlying influences on their distribution throughout diverse human habitats. Two complementary approaches were used: the first employed retrospective analyses of notification data to examine disease patterns across urban and rural environments (Chapters 3 and 4). This identified high risk locations, and enabled some inferences to be made about potential transmission risk factors. The second approach focused on the collection of field and laboratory data to expand knowledge of how specific vector‐host interactions might contribute to arboviral transmission pathways in different environments (Chapters 5 and 6).
In the first approach, I examined recent disease patterns for dengue and Ross River viruses through retrospective data analyses. Each virus had exhibited recent increases in incidence rates, and potential for further expansion of both the magnitude and the geographic distribution of outbreaks. For both dengue and RRV, I aimed to identify some of the spatial, temporal and demographic risk factors for infection, and to explore potential environmental influences on the trends observed. For each virus, I identified specific epidemiological characteristics that were unique to each geographic region studied.
For dengue, the highest burden is often found in urban cities of Asia. However, I found that in Sabah, rates of urban and rural incidence rates were equivalent. Similar
Chapter 7: General discussion 202
or variable urban‐rural trends are also observed in other countries, but are not necessarily well‐documented, nor are the drivers well‐understood [40, 41, 123, 127]. This is also the case for RRV, with the specific drivers of disease trends across diverse Australian, and Pacific Island, habitats not yet identified. My work in Chapter 4 showed that in one endemic region, South East Queensland, the RRV burden is greatest in areas where urban and rural environments intersect. In terms of understanding RRV transmission, this provides one more piece of what is a large jigsaw puzzle.
A common factor influencing the distribution of both viruses is environmental change. The urban and rural distributions of both viruses are likely the result of adaptation to primarily human‐driven changes in the environment that can alter arboviral transmission dynamics. Two of the most significant human‐induced environmental changes known to dramatically disturb natural ecosystems are agricultural development and urbanisation [33, 245]. Each of these have demonstrated potential to increase zoonotic disease risk through land use changes that alter the balance of vectors and hosts present [88, 90, 111, 246]. Although this thesis did not assess environmental change specifically, in both Sabah and Queensland, the observation of high disease incidence rates in regions having undergone substantial environmental change suggests that altered ecosystem dynamics might contribute to the disease distributions of both arboviruses. For example, in Sabah, the expansion of the palm oil plantation industry in rural districts may have enhanced breeding opportunities for Aedes albopictus; while in South East Queensland, the encroachment of residential and agricultural development into wildlife conservation areas may have brought humans into much closer contact with RRV vectors and hosts. Economic development processes such as these that dramatically alter vector, host and human contact may help to explain the spatial patterns that were observed for both viruses.
The demographic and temporal trends identified here for DENV and RRV, including age distributions affected and the timing of seasonal peaks, were consistent with other published findings for each virus, and did not appear to have undergone significant recent change. It could be possible that the influences on demographic
Chapter 7: General discussion 203
change for these two viruses (such as virus evolution, herd immunity etc.) act over a longer‐term period than that studied here. Another potential driver of disease patterns for both dengue and RRV is human movement and behaviour. These are known to play a substantial role in the movement of viruses, vectors and hosts within and between habitats, and to create opportunities for interaction [247, 248]. It was not possible in this PhD to investigate the impacts of human movement and behaviour on disease patterns, though an assessment of their influence would be valuable to include in future studies. Current and future disease trends for both DENV and RRV would also be influenced by additional local and regional factors not assessed here, including fluctuations in climate, human behaviour, the availability of specific vector and host habitats [90, 113, 128, 169, 220], and by virological factors, such as the evolutionary rate and dynamics of virus‐vector‐host adaptation [249]. Each of these will have a strong influence on the timing and distribution of arboviral epidemics.
Many of these influences on DENV and RRV incidence are also drivers of other arboviral and vector‐borne diseases. Therefore, gaining improved knowledge of the determinants of transmission of DENV and RRV may also contribute to understanding and prevention of additional arboviral diseases. Collection of more detailed data on the array of influences on arboviral transmission that operate at different spatial scales could contribute to prevention and control of both current and future arboviral disease threats. With the ongoing processes of population growth, urban development and changing climate conditions, there is a strong likelihood that the burden of dengue, RRV, and other arboviruses will continue to increase [12, 250, 251]. Addressing this challenge requires greater understanding of virus and vector ecology, improved arboviral surveillance, and a well‐developed tool kit of response mechanisms. For example, arboviral surveillance could be expanded to capture more detailed data on arbovirus circulation patterns across different environments, and disease notification systems could capture more information about specific human activities that increase exposure to infection.
The importance of assessing arboviral infection risks across the variety of ecological settings, and at‐risk populations, in which they circulate was highlighted
Chapter 7: General discussion 204
through the findings of Chapters 3 and 4. These contributed to knowledge of dengue and RRV disease dynamics in specific transmission settings, and provided a foundation for the chapters to follow. For RRV in particular, there are major knowledge gaps in evidence implicating specific vectors and hosts in RRV transmission across different environments. In order to address this gap, the research presented in Chapters 5 and 6 focused on a more in‐depth investigation of specific RRV vectors and hosts in South East Queensland. Although a full investigation of the 72 persistent hot spot locations identified in Chapter 4 was outside the scope of this thesis, a pilot study exploring localised transmission dynamics across 4 sites in the endemic city of Brisbane was performed (Chapter 6). To do this, field data on vectors and hosts present in urban and suburban environments of Brisbane were used together with a novel laboratory approach (described in Chapter 5) to enhance the information possible to obtain about RRV transmission.
The micro‐PRNT laboratory technique developed as part of this research is a scaled‐down version of the gold standard technique for viral antibody surveillance, and has previously been applied to high‐throughput screening of vertebrate blood or tissue samples for specific pathogens [252, 253]. I present here the first demonstration that this technique is an effective tool for screening mosquito blood meals. Using this method, in combination with established tools for the origin of the blood meal, a single blood‐fed mosquito will yield information on several components of potential disease transmission pathways including a) the mosquito species b) its vertebrate host and c) exposure of that host to RRV. With the wide range of potential vectors and host species that could be involved in RRV transmission, this approach assists in narrowing down the species that could form the key transmission pathways and that may be the focus of risk assessment and disease mitigation strategies. It also offers significant potential as a new platform for exploring zoonotic transmission pathways.
The application of this approach to blood‐fed mosquito samples collected during field studies conducted in Brisbane resulted in the assessment of RRV antibody presence in mosquito blood meals from known RRV vectors. This linkage of specific vector species, their host feeding choices, and the prior RRV exposure status of those
Chapter 7: General discussion 205
hosts, enabled insight into potential RRV transmission pathways. In addition, this study added to the limited literature on blood feeding preferences of mosquitoes in urban areas. The results were somewhat surprising in terms of the large proportion of feeding on humans and birds rather than other mammals or marsupials which were previously thought to have been important in the RRV transmission cycle. This finding suggests that the contribution of humans to mediating ongoing transmission should be reconsidered. Particularly in urban areas, it is possible that humans play a greater role than previously thought. This hypothesis is supported by the aforementioned Pacific Island epidemics on the early 1980s, and the ongoing circulation of RRV outside of the endemic range of marsupial reservoirs, which further suggest possible human involvement.
Although the studies of Chapters 5 and 6 do not conclusively implicate any one vector or host in RRV transmission, they provide a basis for scaling up future studies. Ideally, these studies should also include viral isolation, since blood feeding patterns and the antibody status of wildlife alone do not indicate the relative roles and competence of different hosts. However, virus‐specific antibodies are a far easier target in studies of host and disease associations than the virus itself. This is because most vertebrate reservoirs have short viraemic periods and the probability of capturing a blood‐fed mosquito with a viraemic blood meal are low. With expanded investigation, recurrent patterns of vector, host, and antibody associations will yield new clues about RRV transmission pathways in urban and suburban areas. With the work performed here, I provided a convincing proof of principle for use of a novel combination of methods that, if expanded in scale, could help define important specific vector and host species for characterising transmission, in specific environments and in relation to seasonal changes. The use of a research design encompassing collection of wildlife, entomology and molecular diagnostic data enabled more informative results than any individual aspect might have yielded.
Overall, my research has contributed to the understanding of both dengue and RRV disease patterns and potential demographic and ecological influences on transmission. As environments are continually modified by humans, it will be increasingly important to monitor how this affects the spill‐over of zoonotic disease
Chapter 7: General discussion 206
to the human population, and to proactively explore how this might be managed. This thesis helps to highlight the potential for continuing arboviral range expansion and illustrates how deployment of the full range of surveillance, ecological and xenodiagnostic tools is required to describe and predict that expansion. I provided a methodological framework for exploring the linkages between human disease patterns and the ecological balance of arboviral vectors and hosts in different transmission settings. My work illustrates the need to employ new and diverse approaches to answering essential public health questions, and to developing solutions to future disease problems.
Future research of arboviral ecology should ideally incorporate more detailed surveillance of human, virus and vector circulation across different endemic settings; for example, investigating whether human, virus or vector movement differ between rural and urban areas. More detailed investigation of demographic risk factors of infection may help to target preventive actions by identifying potential occupational groups at risk (such as occupations based indoors versus outdoors), and/or human behaviours that increase mosquito biting risk (e.g. water storage and waste disposal practices, knowledge and attitudes toward personal prevention strategies). Assessing additional demographic data may also provide more insight into the environmental characteristics associated with infection, and the relative involvement of specific vectors and hosts, although vector and host incrimination studies merit separate investigation.
For dengue, reducing vector‐human interactions is particularly important, but it relies upon having adequate knowledge of vector biology to design effective preventative strategies, which will require additional investments in research. Current innovations include infecting Ae. aegypti with Wolbachia bacteria to decrease their capacity to transmit viruses [254, 255], and use of genetic techniques to induce sterility or block transmission in vector species, reducing their infective populations [256, 257]. However, many of these techniques are not yet at a stage where they are quick and easy to implement, and at large scales. Sustainable reductions in the rapid spread of dengue and other arboviruses will likely require a combined application, or optimisation, of existing and new tools [258, 259]. In
Chapter 7: General discussion 207
addition, enhanced arbovirus and vector surveillance to enable prediction of and response to new arboviral threats is necessary, especially in tropical regions where many arboviruses circulate [260].
Compared with dengue, RRV research and development of control methods are comparatively less. No specific surveillance or control manuals exist for RRV while for dengue there are both global and national strategies in place [261]. While this is arguably justified given RRV’s lower overall health burden compared with dengue, RRV exerts substantial ongoing health and socioeconomic impacts in Australia, and in neighbouring countries, although in the latter are poorly quantified. In Australia alone, it was estimated that individual treatment costs were at least AU$5.5 million per year during 2006‐2015, with at least AU$10 million spent annually for public health education, mosquito control and research [242]. While some research investment has been made in developing a vaccine and clinical therapies for RRV, research effort to address the lack of adequate surveillance and prevention strategies for RRV has been neglected.
For Australia in particular, greater investment in research and preventive approaches for RRV are warranted, and would likely have flow on benefits for mitigating additional current and emerging arboviruses, such as Barmah Forest virus, Murray Valley encephalitis virus, and other pathogens that share the same vectors [47, 262‐265]. But the impacts of RRV are not limited only to Australia – RRV has shown potential to expand its current geographic range to infect new populations, as evidenced by the large epidemic throughout the South Pacific region in the early 1980s, as well as the recent discovery of RRV’s silent expansion throughout additional Pacific Island countries [55, 56, 59]. This additional evidence of RRVs capacity to extend its geographic distribution underscores the need for improved preparedness and responsiveness, should RRV become a global threat.
RRV prevention requires significantly more research investment before specific control strategies can be implemented. At a minimum, the specific vectors associated with regular epidemics around Australia must be identified. This requires timely deployment of mosquito and virus surveillance tools in conjunction with the known seasonal peak in transmission. Ideally, future studies would also build a broader
Chapter 7: General discussion 208
understanding of RRV transmission ecology. In particular, understanding how and when vector‐host‐human interactions at urban‐rural interfaces could result in human spillover. This information could then be used to design mitigation strategies. Gaining a better understanding of vector and host ecology at a local level could not only help address the current and future burdens of RRV disease, but may also contribute to preparedness for future emerging arboviral disease threats.
Both dengue and RRV would benefit from enhanced surveillance activities, including case follow‐up (data on onset dates, travel history, primary and secondary places of exposure), linked to surveillance of circulating viruses and to data on vector distributions. This would aid both identification of likely location(s) of infection and foci for response activities, and build understanding of the role of key risk factors, such as human movement and behaviour, in the spread of epidemics. These surveillance activities could form integrated notification systems for broad arboviral disease risks, and could guide more cost‐effective deployment of mosquito control activities. Similarly, an integrated system could inform timely public health messages around seasonal risk, and the strategies to be implemented in response. In order to address the impacts of current and emerging arboviral diseases, it is essential to invest in surveillance programs and the methods that support them. This will require new approaches and a multidisciplinary response, giving greater consideration to the effects of environmental change and urban development on human health, and to ameliorating their negative effects.
More broadly, in addition to improved surveillance strategies, arboviral disease prevention and control would be strengthened by closer monitoring and management of environmental change, and stronger linkage between government departments collecting human demographic and case reporting data, vector surveillance data, virus and serological surveillance data, and environmental and wildlife survey data. Outbreak responses will still rely on the availability of practical and effective control tools in endemic communities, which will vary according to the public health resources available in different countries. Given that changing environments have provided an opportunity for a number of arboviral and zoonotic diseases to emerge, investment in development of an integrated combination of both
Chapter 7: General discussion 209
targeted and generalised biological, environmental and vector control interventions that can potentially be directed toward multiple arboviral diseases could be a sustainable approach [11, 260]. These could include improved environmental management and urban planning strategies, along with enhanced surveillance and response. Building multisectoral cooperation in these areas would provide a foundation to enable more timely and targeted public health education campaigns, more efficient deployment of vector control resources, and more accurate prediction and management of epidemics.
Arboviral disease can be influenced by a broad range of factors including weather patterns, social and economic factors, and rapidly changing ecological processes across diverse environments. Investigating and quantifying the relative impact of these on disease burdens can be scientifically and logistically challenging, but is necessary to identify the critical points where intervention is possible. Without adequately addressing knowledge gaps on the specific drivers and pathways of emerging arboviral diseases, it will be impossible to formulate prevention or mitigation strategies that identify and target specific elements of the transmission pathway. Knowledge of those factors would inform intervention programs that might include vector‐specific control measures, wildlife and environmental management, urban planning (e.g. reducing development close to mosquito habitat, or facilitating better waste disposal in urban areas) or education campaigns to identify risky behaviours and reduce mosquito‐human contact. Filling the knowledge gaps requires financial investment and multidisciplinary collaboration but the tremendous human health and economic costs of these diseases, and the threat posed by as yet unknown arboviral threats, justifies that expense.
In conclusion, this thesis suggests collecting more precise information on arboviral infection risk factors in support of preparedness for future disease emergence. This would enable more accurate prediction, and therefore prevention, of epidemics through development of heuristic models, early warning systems, and strong linkage of data across sectors. These should be informed by studies to improve understanding of virus and vector dissemination pathways, incriminate specific vectors and hosts in specific environments, and evaluate the most appropriate
Chapter 7: General discussion 210
targets of control measures. The rural‐urban transition is a key implication here and needs to be better studied, as clearly rural areas are also major foci of disease and play important roles in virus transmission. With arboviral disease burdens predicted to further rise and expand, proactive efforts to clarify, monitor and respond to the underlying drivers are essential.
Chapter 7: General discussion 211
Chapter 8: References
1. Vector‐borne diseases fact sheet. World Health Organization, Geneva. October 31, 2017 [cited January 4, 2020]; Available from: https://www.who.int/news‐ room/fact‐sheets/detail/vector‐borne‐diseases. 2. Global vector control response 2017–2030. World Health Organization, Geneva; 2017. Licence: CC BY‐NC‐SA 3.0 IGO. 3. A global brief on vector‐borne diseases. World Health Organization, Geneva, Switzerland; 2014. 4. Gubler, D.J., The global emergence/resurgence of arboviral diseases as public health problems. Arch Med Res, 2002. 33(4): p. 330‐42. 5. Musso, D., et al., Unexpected outbreaks of arbovirus infections: lessons learned from the Pacific and tropical America. Lancet Infect Dis, 2018. 18(11): p. e355‐e361. 6. Kilpatrick, A.M. and S.E. Randolph, Drivers, dynamics, and control of emerging vector‐borne zoonotic diseases. Lancet, 2012. 380(9857): p. 1946‐55. 7. Harrus, S. and G. Baneth, Drivers for the emergence and re‐emergence of vector‐borne protozoal and bacterial diseases. Int J Parasitol, 2005. 35(11‐12): p. 1309‐18. 8. Mayer, S.V., R.B. Tesh, and N. Vasilakis, The emergence of arthropod‐borne viral diseases: A global prospective on dengue, chikungunya and zika fevers. Acta Trop, 2017. 166: p. 155‐163. 9. Wilder‐Smith, A. and T.P. Monath, Responding to the threat of urban yellow fever outbreaks. Lancet Infect Dis, 2017. 17(3): p. 248‐250. 10. Gould, E., et al., Emerging arboviruses: Why today? One Health, 2017. 4: p. 1‐ 13. 11. Wilder‐Smith, A., et al., Epidemic arboviral diseases: priorities for research and public health. Lancet Infect Dis, 2017. 17(3): p. e101‐e106. 12. Weaver, S.C. and W.K. Reisen, Present and future arboviral threats. Antiviral Res, 2010. 85(2): p. 328‐45. 13. Crawford, J.E., et al., Population genomics reveals that an anthropophilic population of Aedes aegypti mosquitoes in West Africa recently gave rise to American and Asian populations of this major disease vector. BMC Biol, 2017. 15(1): p. 16. 14. Kraemer, M.U., et al., The global distribution of the arbovirus vectors Aedes aegypti and Ae. albopictus. Elife, 2015. 4: p. e08347. 15. Maynard, A.J., et al., Tiger on the prowl: Invasion history and spatio‐temporal genetic structure of the Asian tiger mosquito Aedes albopictus (Skuse 1894) in the Indo‐Pacific. PLoS Negl Trop Dis, 2017. 11(4): p. e0005546. 16. Hawley, W.A., et al., Aedes albopictus in North America: probable introduction in used tires from northern Asia. Science, 1987. 236(4805): p. 1114‐6. 17. Powell, J.R., Mosquitoes on the move. Science, 2016. 354(6315): p. 971‐972.
Chapter 8: References 212
18. Bhatt, S., et al., The global distribution and burden of dengue. Nature, 2013. 496(7446): p. 504‐7. 19. Dengue and severe dengue factsheet. World Health Organization; Geneva, Switzerland. [cited [Accessed January 5, 2020; Available from: https://www.who.int/news‐room/fact‐sheets/detail/dengue‐and‐severe‐ dengue. 20. Kyle, J.L. and E. Harris, Global spread and persistence of dengue. Annu Rev Microbiol, 2008. 62: p. 71‐92. 21. Rezza, G., et al., Infection with chikungunya virus in Italy: an outbreak in a temperate region. Lancet, 2007. 370(9602): p. 1840‐6. 22. Riccardo, F., et al., Secondary Autochthonous Outbreak of Chikungunya, Southern Italy, 2017. Emerg Infect Dis, 2019. 25(11): p. 2093‐2095. 23. Weaver, S.C., et al., Zika, Chikungunya, and Other Emerging Vector‐Borne Viral Diseases. Annu Rev Med, 2018. 69: p. 395‐408. 24. Leparc‐Goffart, I., et al., Chikungunya in the Americas. Lancet, 2014. 383(9916): p. 514. 25. Lessler, J., et al., Assessing the global threat from Zika virus. Science, 2016. 353(6300): p. aaf8160. 26. Yellow fever fact sheet. World Health Organization, Geneva, Switzerland. [cited [Accessed January 4, 2020; Available from: https://www.who.int/news‐ room/fact‐sheets/detail/yellow‐fever. 27. Wang, L., et al., Yellow fever virus: Increasing imported cases in China. J Infect, 2016. 73(4): p. 377‐80. 28. Leisnham, P.T. and S.A. Juliano, Impacts of climate, land use, and biological invasion on the ecology of immature Aedes mosquitoes: implications for La Crosse emergence. Ecohealth, 2012. 9(2): p. 217‐28. 29. Reiter, P., Climate change and mosquito‐borne disease: knowing the horse before hitching the cart. Rev Sci Tech, 2008. 27(2): p. 383‐98. 30. Franklinos, L.H.V., et al., The effect of global change on mosquito‐borne disease. Lancet Infect Dis, 2019. 19(9): p. e302‐e312. 31. Wartel, T.A., et al., Three Decades of Dengue Surveillance in Five Highly Endemic South East Asian Countries. Asia Pac J Public Health, 2017. 29(1): p. 7‐ 16. 32. Reiner, R.C., Jr., et al., Quantifying the Epidemiological Impact of Vector Control on Dengue. PLoS Negl Trop Dis, 2016. 10(5): p. e0004588. 33. Gubler, D.J., Dengue, Urbanization and Globalization: The Unholy Trinity of the 21(st) Century. Trop Med Health, 2011. 39(4 Suppl): p. 3‐11. 34. Gubler, D.J., Epidemic dengue/dengue hemorrhagic fever as a public health, social and economic problem in the 21st century. Trends Microbiol, 2002. 10(2): p. 100‐3. 35. Hamel, R., et al., Phylogenetic analysis revealed the co‐circulation of four dengue virus serotypes in Southern Thailand. PLoS One, 2019. 14(8): p. e0221179. 36. Galarion, M.J., et al., Genotypic persistence of dengue virus in the Philippines. Infect Genet Evol, 2019. 69: p. 134‐141.
Chapter 8: References 213
37. Gintarong, T.J., et al., Circulation of all dengue virus serotypes during dengue outbreak in Sandakan, Sabah, Malaysia (2016). J Vector Borne Dis, 2018. 55(2): p. 168‐171. 38. Vasilakis, N., et al., Fever from the forest: prospects for the continued emergence of sylvatic dengue virus and its impact on public health. Nat Rev Microbiol, 2011. 9(7): p. 532‐41. 39. Gratz, N.G., Critical review of the vector status of Aedes albopictus. Med Vet Entomol, 2004. 18(3): p. 215‐27. 40. Yoon, I.K., et al., Fine scale spatiotemporal clustering of dengue virus transmission in children and Aedes aegypti in rural Thai villages. PLoS Negl Trop Dis, 2012. 6(7): p. e1730. 41. Vannavong, N., et al., Effects of socio‐demographic characteristics and household water management on Aedes aegypti production in suburban and rural villages in Laos and Thailand. Parasit Vectors, 2017. 10(1): p. 170. 42. Rohani, A., et al., Eco‐virological survey of Aedes mosquito larvae in selected dengue outbreak areas in Malaysia. J Vector Borne Dis, 2014. 51(4): p. 327‐32. 43. Rezza, G., Aedes albopictus and the reemergence of Dengue. BMC Public Health, 2012. 12: p. 72. 44. Johari, N.A., et al., Sylvatic dengue virus type 4 in Aedes aegypti and Aedes albopictus mosquitoes in an urban setting in Peninsular Malaysia. PLoS Negl Trop Dis, 2019. 13(11): p. e0007889. 45. Claflin, S.B. and C.E. Webb, Ross River Virus: Many Vectors and Unusual Hosts Make for an Unpredictable Pathogen. PLoS Pathog, 2015. 11(9): p. e1005070. 46. National Notifiable Diseases Surveillance System. Australian Government. [Accessed September 23, 2018]; Available from: http://www9.health.gov.au/cda/source/cda‐index.cfm. 47. van den Hurk, A.F., et al., Emerging tropical diseases in Australia. Part 4. Mosquitoborne diseases. Ann Trop Med Parasitol, 2010. 104(8): p. 623‐40. 48. Amin, J., et al., Ross River virus infection in the north‐west outskirts of the Sydney basin. Commun Dis Intell Q Rep, 1998. 22(6). 49. Russell, R.C., Ross River virus: ecology and distribution. Annu Rev Entomol, 2002. 47: p. 1‐31. 50. Increased incidence and expanded geographical range of Ross River virus. Department of Health and Human Services Report: Victoria. July 2017. 51. Jansen, C.C., et al., Epidemiologic, Entomologic, and Virologic Factors of the 2014–15 Ross River Virus Outbreak, Queensland, Australia. Emerg Infect Dis, 2019. 25(12). 52. Lindsay, M., et al., An outbreak of Ross River virus disease in Southwestern Australia. Emerg Infect Dis, 1996. 2(2): p. 117‐20. 53. Stephenson, E.B., et al., The non‐human reservoirs of Ross River virus: a systematic review of the evidence. Parasit Vectors, 2018. 11(1): p. 188. 54. Harley, D., A. Sleigh, and S. Ritchie, Ross River virus transmission, infection, and disease: a cross‐disciplinary review. Clin Microbiol Rev, 2001. 14(4): p. 909‐32, table of contents. 55. Aubry, M., et al., Silent Circulation of Ross River Virus in French Polynesia. Int J Infect Dis, 2015. 37: p. 19‐24.
Chapter 8: References 214
56. Lau, C., et al., New evidence for endemic circulation of Ross River virus in the Pacific Islands and the potential for emergence. Int J Infect Dis, 2017. 57: p. 73‐76. 57. Aubry, M., et al., Ross River Virus Antibody Prevalence, Fiji Islands, 2013‐2015. Emerg Infect Dis, 2019. 25(4): p. 827‐830. 58. Aubry, M., et al., Ross River Virus Seroprevalence, French Polynesia, 2014‐ 2015. Emerg Infect Dis, 2017. 23(10): p. 1751‐1753. 59. Shanks, G.D., Could Ross River Virus Be the Next Zika? J Travel Med, 2019. 60. Flies, E.J., et al., Another Emerging Mosquito‐Borne Disease? Endemic Ross River Virus Transmission in the Absence of Marsupial Reservoirs. BioScience, 2018. 68(4): p. 288‐293. 61. Mackenzie, J.S., D.J. Gubler, and L.R. Petersen, Emerging flaviviruses: the spread and resurgence of Japanese encephalitis, West Nile and dengue viruses. Nat Med, 2004. 10(12 Suppl): p. S98‐109. 62. Weaver, S.C. and N.L. Forrester, Chikungunya: Evolutionary history and recent epidemic spread. Antiviral Res, 2015. 120: p. 32‐9. 63. Shepard, D.S., et al., The global economic burden of dengue: a systematic analysis. Lancet Infect Dis, 2016. 16(8): p. 935‐41. 64. Pfeffer, M. and G. Dobler, Emergence of zoonotic arboviruses by animal trade and migration. Parasit Vectors, 2010. 3(1): p. 35. 65. Gubler, D.J., The global resurgence of arboviral diseases. Trans R Soc Trop Med Hyg, 1996. 90(5): p. 449‐51. 66. Marklewitz, M. and S. Junglen, Evolutionary and ecological insights into the emergence of arthropod‐borne viruses. Acta Trop, 2019. 190: p. 52‐58. 67. Liang, G., X. Gao, and E.A. Gould, Factors responsible for the emergence of arboviruses; strategies, challenges and limitations for their control. Emerg Microbes Infect, 2015. 4(3): p. e18. 68. Gubler, D.J., Cities spawn epidemic dengue viruses. Nat Med, 2004. 10(2): p. 129‐30. 69. Gubler, D.J., Dengue and dengue hemorrhagic fever. Clin Microbiol Rev, 1998. 11(3): p. 480‐96. 70. Brathwaite Dick, O., et al., The history of dengue outbreaks in the Americas. Am J Trop Med Hyg, 2012. 87(4): p. 584‐93. 71. Hayes, E.B. and D.J. Gubler, West Nile virus: epidemiology and clinical features of an emerging epidemic in the United States. Annu Rev Med, 2006. 57: p. 181‐94. 72. West Nile virus factsheet. World Health Organization; Geneva, Switzerland. [cited [Accessed January 7, 2020; Available from: https://www.who.int/news‐ room/fact‐sheets/detail/west‐nile‐virus. 73. Kilpatrick, A.M., Globalization, land use, and the invasion of West Nile virus. Science, 2011. 334(6054): p. 323‐7. 74. Musso, D. and D.J. Gubler, Zika Virus. Clin Microbiol Rev, 2016. 29(3): p. 487‐ 524. 75. Baud, D., et al., Zika virus: A new threat to human reproduction. Am J Reprod Immunol, 2017. 77(2).
Chapter 8: References 215
76. Bryant, J.E., E.C. Holmes, and A.D. Barrett, Out of Africa: a molecular perspective on the introduction of yellow fever virus into the Americas. PLoS Pathog, 2007. 3(5): p. e75. 77. Gossner, C.M., et al., Increased risk of yellow fever infections among unvaccinated European travellers due to ongoing outbreak in Brazil, July 2017 to March 2018. Euro Surveill, 2018. 23(11). 78. Rezza, G., Chikungunya is back in Italy: 2007‐2017. J Travel Med, 2018. 25(1). 79. Sudeep, A.B. and D. Parashar, Chikungunya: an overview. J Biosci, 2008. 33(4): p. 443‐9. 80. Dupuis‐Maguiraga, L., et al., Chikungunya disease: infection‐associated markers from the acute to the chronic phase of arbovirus‐induced arthralgia. PLoS Negl Trop Dis, 2012. 6(3): p. e1446. 81. Abreu, F.V.S., et al., Haemagogus leucocelaenus and Haemagogus janthinomys are the primary vectors in the major yellow fever outbreak in Brazil, 2016‐2018. Emerg Microbes Infect, 2019. 8(1): p. 218‐231. 82. Monath, T.P., Facing up to re‐emergence of urban yellow fever. Lancet, 1999. 353(9164): p. 1541. 83. Moreira‐Soto, A., et al., Evidence for multiple sylvatic transmission cycles during the 2016‐2017 yellow fever virus outbreak, Brazil. Clin Microbiol Infect, 2018. 24(9): p. 1019 e1‐1019 e4. 84. Moussallem, T.M., et al., Yellow fever outbreak in a rural‐urban mixed community of Espirito Santo, Brazil: epidemiological aspects. Rev Panam Salud Publica, 2019. 43: p. e29. 85. Young, K.I., et al., Abundance and distribution of sylvatic dengue virus vectors in three different land cover types in Sarawak, Malaysian Borneo. Parasit Vectors, 2017. 10(1): p. 406. 86. Pereira Dos Santos, T., et al., Potential of Aedes albopictus as a bridge vector for enzootic pathogens at the urban‐forest interface in Brazil. Emerg Microbes Infect, 2018. 7(1): p. 191. 87. Ramasamy, R. and S.N. Surendran, Mosquito vectors developing in atypical anthropogenic habitats: Global overview of recent observations, mechanisms and impact on disease transmission. J Vector Borne Dis, 2016. 53(2): p. 91‐8. 88. Bradley, C.A., S.E. Gibbs, and S. Altizer, Urban land use predicts West Nile virus exposure in songbirds. Ecol Appl, 2008. 18(5): p. 1083‐92. 89. Zahouli, J.B.Z., et al., Effect of land‐use changes on the abundance, distribution, and host‐seeking behavior of Aedes arbovirus vectors in oil palm‐ dominated landscapes, southeastern Cote d'Ivoire. PLoS One, 2017. 12(12): p. e0189082. 90. Patz, J.A., et al., Disease emergence from global climate and land use change. Med Clin North Am, 2008. 92(6): p. 1473‐91, xii. 91. Viral infections and global change, ed. S.K. Singh. 2014, Hoboken, New Jersey: Wiley. 92. Wang, E., et al., Evolutionary relationships of endemic/epidemic and sylvatic dengue viruses. J Virol, 2000. 74(7): p. 3227‐34. 93. Simmons, C.P., et al., Dengue. N Engl J Med, 2012. 366(15): p. 1423‐32. 94. Guzman, M.G. and E. Harris, Dengue. Lancet, 2015. 385(9966): p. 453‐65.
Chapter 8: References 216
95. Silva, M.M.O., et al., Concomitant Transmission of Dengue, Chikungunya, and Zika Viruses in Brazil: Clinical and Epidemiological Findings From Surveillance for Acute Febrile Illness. Clin Infect Dis, 2019. 69(8): p. 1353‐1359. 96. Capeding, M.R., et al., Dengue and other common causes of acute febrile illness in Asia: an active surveillance study in children. PLoS Negl Trop Dis, 2013. 7(7): p. e2331. 97. Vasilakis, N. and S.C. Weaver, The history and evolution of human dengue emergence. Adv Virus Res, 2008. 72: p. 1‐76. 98. Godding, C.C., An Account of an Obscure Outbreak of Dengue Occurring on Board H.M.S. Agamemnon while Stationed at Zanzibar between November, 1888, and September, 1889. Br Med J, 1890. 1(1520): p. 352‐4. 99. Smith, C.E., The history of dengue in tropical Asia and its probable relationship to the mosquito Aedes aegypti. J Trop Med Hyg, 1956. 59(10): p. 243‐51. 100. Gilbertson, W.E., Sanitary Aspects of the Control of the 1943‐1944 Epidemic of Dengue Fever in Honolulu. Am J Public Health Nations Health, 1945. 35(3): p. 261‐70. 101. Hotta, S., Experimental studies on dengue. I. Isolation, identification and modification of the virus. J Infect Dis, 1952. 90(1): p. 1‐9. 102. Hammon, W.M., A. Rudnick, and G.E. Sather, Viruses associated with epidemic hemorrhagic fevers of the Philippines and Thailand. Science, 1960. 131(3407): p. 1102‐3. 103. Messina, J.P., et al., Global spread of dengue virus types: mapping the 70 year history. Trends Microbiol, 2014. 22(3): p. 138‐46. 104. Williams, M., et al., Lineage II of Southeast Asian/American DENV‐2 is associated with a severe dengue outbreak in the Peruvian Amazon. Am J Trop Med Hyg, 2014. 91(3): p. 611‐20. 105. Nisalak, A., et al., Serotype‐specific dengue virus circulation and dengue disease in Bangkok, Thailand from 1973 to 1999. Am J Trop Med Hyg, 2003. 68(2): p. 191‐202. 106. Katzelnick, L.C., et al., Antibody‐dependent enhancement of severe dengue disease in humans. Science, 2017. 358(6365): p. 929‐932. 107. Muzari, M.O., et al., Dominance of the tiger: The displacement of Aedes aegypti by Aedes albopictus in parts of the Torres Strait, Australia. Commun Dis Intell (2018), 2019. 43. 108. Hashim, N.A., et al., Co‐breeding Association of Aedes albopictus (Skuse) and Aedes aegypti (Linnaeus) (Diptera: Culicidae) in Relation to Location and Container Size. Trop Life Sci Res, 2018. 29(1): p. 213‐227. 109. Braks, M.A., et al., Convergent habitat segregation of Aedes aegypti and Aedes albopictus (Diptera: Culicidae) in southeastern Brazil and Florida. J Med Entomol, 2003. 40(6): p. 785‐94. 110. Ferreira‐de‐Lima, V.H. and T.N. Lima‐Camara, Natural vertical transmission of dengue virus in Aedes aegypti and Aedes albopictus: a systematic review. Parasit Vectors, 2018. 11(1): p. 77. 111. Li, Y., et al., Urbanization increases Aedes albopictus larval habitats and accelerates mosquito development and survivorship. PLoS Negl Trop Dis, 2014. 8(11): p. e3301.
Chapter 8: References 217
112. Gottdenker, N.L., et al., Anthropogenic land use change and infectious diseases: a review of the evidence. Ecohealth, 2014. 11(4): p. 619‐32. 113. Astrom, C., et al., Potential distribution of dengue fever under scenarios of climate change and economic development. Ecohealth, 2012. 9(4): p. 448‐54. 114. Clyde, K., J.L. Kyle, and E. Harris, Recent advances in deciphering viral and host determinants of dengue virus replication and pathogenesis. J Virol, 2006. 80(23): p. 11418‐31. 115. Parham, P.E., et al., Climate, environmental and socio‐economic change: weighing up the balance in vector‐borne disease transmission. Philos Trans R Soc Lond B Biol Sci, 2015. 370(1665). 116. Patz, J.A. and S.H. Olson, Climate change and health: global to local influences on disease risk. Ann Trop Med Parasitol, 2006. 100(5‐6): p. 535‐49. 117. Teurlai, M., et al., Socio‐economic and Climate Factors Associated with Dengue Fever Spatial Heterogeneity: A Worked Example in New Caledonia. PLoS Negl Trop Dis, 2015. 9(12): p. e0004211. 118. Descloux, E., et al., Climate‐based models for understanding and forecasting dengue epidemics. PLoS Negl Trop Dis, 2012. 6(2): p. e1470. 119. Brady, O.J., et al., Global temperature constraints on Aedes aegypti and Ae. albopictus persistence and competence for dengue virus transmission. Parasit Vectors, 2014. 7: p. 338. 120. Cheng, Q., et al., Climate and the Timing of Imported Cases as Determinants of the Dengue Outbreak in Guangzhou, 2014: Evidence from a Mathematical Model. PLoS Negl Trop Dis, 2016. 10(2): p. e0004417. 121. Hales, S., et al., Potential effect of population and climate changes on global distribution of dengue fever: an empirical model. Lancet, 2002. 360(9336): p. 830‐4. 122. Williams, C.R., et al., Testing the impact of virus importation rates and future climate change on dengue activity in Malaysia using a mechanistic entomology and disease model. Epidemiol Infect, 2015. 143(13): p. 2856‐64. 123. Dubot‐Peres, A., et al., An epidemic of dengue‐1 in a remote village in rural Laos. PLoS Negl Trop Dis, 2013. 7(8): p. e2360. 124. Naish, S., et al., Climate change and dengue: a critical and systematic review of quantitative modelling approaches. BMC Infect Dis, 2014. 14: p. 167. 125. Misslin, R., et al., Urban climate versus global climate change‐what makes the difference for dengue? Ann N Y Acad Sci, 2016. 1382(1): p. 56‐72. 126. Thai, K.T. and K.L. Anders, The role of climate variability and change in the transmission dynamics and geographic distribution of dengue. Exp Biol Med (Maywood), 2011. 236(8): p. 944‐54. 127. Teurlai, M., et al., Can human movements explain heterogeneous propagation of dengue fever in Cambodia? PLoS Negl Trop Dis, 2012. 6(12): p. e1957. 128. Akter, R., et al., Joint effects of climate variability and socioecological factors on dengue transmission: epidemiological evidence. Trop Med Int Health, 2017. 22(6): p. 656‐669. 129. Stewart Ibarra, A.M., et al., Dengue vector dynamics (Aedes aegypti) influenced by climate and social factors in Ecuador: implications for targeted control. PLoS One, 2013. 8(11): p. e78263.
Chapter 8: References 218
130. Ebi, K.L. and J. Nealon, Dengue in a changing climate. Environ Res, 2016. 151: p. 115‐123. 131. Halstead, S.B., Mosquito‐borne haemorrhagic fevers of South and South‐East Asia. Bull World Health Organ, 1966. 35(1): p. 3‐15. 132. Mackenzie, J.S. and D.T. Williams, The zoonotic flaviviruses of southern, south‐ eastern and eastern Asia, and Australasia: the potential for emergent viruses. Zoonoses Public Health, 2009. 56(6‐7): p. 338‐56. 133. Ferguson, N.M., Challenges and opportunities in controlling mosquito‐borne infections. Nature, 2018. 559(7715): p. 490‐497. 134. Powers, A.M., et al., Evolutionary relationships and systematics of the alphaviruses. J Virol, 2001. 75(21): p. 10118‐31. 135. Suhrbier, A., M.C. Jaffar‐Bandjee, and P. Gasque, Arthritogenic alphaviruses‐‐ an overview. Nat Rev Rheumatol, 2012. 8(7): p. 420‐9. 136. Doherty, R.L., et al., The isolation of a third group A arbovirus in Australia with preliminary observations on its relationship to epidemic polyarthritis. Aust. J. Sci., 1963. 26: p. 183‐184. 137. Faragher, S.G., et al., Genome sequences of a mouse‐avirulent and a mouse‐ virulent strain of Ross River virus. Virology, 1988. 163(2): p. 509‐26. 138. Sammels, L.M., et al., Geographic distribution and evolution of Ross River virus in Australia and the Pacific Islands. Virology, 1995. 212(1): p. 20‐9. 139. Jones, A., et al., Molecular evolutionary dynamics of Ross River virus and implications for vaccine efficacy. J Gen Virol, 2010. 91(Pt 1): p. 182‐8. 140. Lindsay, M.D., R.J. Coelen, and J.S. Mackenzie, Genetic heterogeneity among isolates of Ross River virus from different geographical regions. J Virol, 1993. 67(6): p. 3576‐85. 141. Michie, A., et al., Genome‐Scale Phylogeny and Evolutionary Analysis of Ross River Virus Reveals Periodic Sweeps of Lineage Dominance in Western Australia, 1977‐2014. J Virol, 2020. 94(2). 142. Mylonas, A.D., et al., Natural history of Ross River virus‐induced epidemic polyarthritis. Med J Aust, 2002. 177(7): p. 356‐60. 143. Condon, R.J. and I.L. Rouse, Acute symptoms and sequelae of Ross River virus infection in South‐Western Australia: a follow‐up study. Clin Diagn Virol, 1995. 3(3): p. 273‐84. 144. Harley, D., et al., Ross River virus disease in tropical Queensland: evolution of rheumatic manifestations in an inception cohort followed for six months. Med J Aust, 2002. 177(7): p. 352‐5. 145. Suhrbier, A. and M. La Linn, Clinical and pathologic aspects of arthritis due to Ross River virus and other alphaviruses. Curr Opin Rheumatol, 2004. 16(4): p. 374‐9. 146. Barber, B., J.T. Denholm, and D. Spelman, Ross River virus. Aust Fam Physician, 2009. 38(8): p. 586‐9. 147. Kelly‐Hope, L.A., D.M. Purdie, and B.H. Kay, Ross River virus disease in Australia, 1886‐1998, with analysis of risk factors associated with outbreaks. J Med Entomol, 2004. 41(2): p. 133‐50. 148. Knope, K.E., et al., Arboviral diseases and malaria in Australia, 2012‐13: Annual report of the National Arbovirus and Malaria Advisory Committee. Commun Dis Intell Q Rep, 2016. 40(1): p. E17‐47.
Chapter 8: References 219
149. Aichinger, G., et al., Safety and immunogenicity of an inactivated whole virus Vero cell‐derived Ross River virus vaccine: a randomized trial. Vaccine, 2011. 29(50): p. 9376‐84. 150. Wressnigg, N., et al., An inactivated Ross River virus vaccine is well tolerated and immunogenic in an adult population in a randomized phase 3 trial. Clin Vaccine Immunol, 2015. 22(3): p. 267‐73. 151. Jupille, H.J., et al., Mutations in nsP1 and PE2 are critical determinants of Ross River virus‐induced musculoskeletal inflammatory disease in a mouse model. Virology, 2011. 410(1): p. 216‐27. 152. Nimmo, J., An unusual epidemic. Med J Aust, 1928. 1: p. 549‐50. 153. Tai, K.S., et al., An outbreak of epidemic polyarthritis (Ross River virus disease) in the Northern Territory during the 1990‐1991 wet season. Med J Aust, 1993. 158(8): p. 522‐5. 154. Jacups, S.P., P.I. Whelan, and B.J. Currie, Ross River virus and Barmah Forest virus infections: a review of history, ecology, and predictive models, with implications for tropical northern Australia. Vector Borne Zoonotic Dis, 2008. 8(2): p. 283‐97. 155. Tesh, R.B., et al., Ross River virus (Togaviridae: Alphavirus) infection (epidemic polyarthritis) in American Samoa. Trans R Soc Trop Med Hyg, 1981. 75(3): p. 426‐31. 156. Rosen, L., D.J. Gubler, and P.H. Bennett, Epidemic polyarthritis (Ross River) virus infection in the Cook Islands. Am J Trop Med Hyg, 1981. 30(6): p. 1294‐ 302. 157. Aaskov, J.G., et al., An epidemic of Ross River virus infection in Fiji, 1979. Am J Trop Med Hyg, 1981. 30(5): p. 1053‐9. 158. Scrimgeour, E.M., J.G. Aaskov, and L.R. Matz, Ross River virus arthritis in Papua New Guinea. Trans R Soc Trop Med Hyg, 1987. 81(5): p. 833‐4. 159. Tesh, R.B., et al., The distribution and prevalence of group A arbovirus neutralizing antibodies among human populations in Southeast Asia and the Pacific islands. Am J Trop Med Hyg, 1975. 24(4): p. 664‐75. 160. Hii, J., et al., Health impact assessments of malaria and Ross River virus infection in the Southern Highlands Province of Papua New Guinea. P N G Med J, 1997. 40(1): p. 14‐25. 161. Jonduo, M.H., G. Bande, and P.F. Horwood, Arboviruses of human health significance in Papua New Guinea. P N G Med J, 2012. 55(1‐4): p. 35‐44. 162. Lau, C., P. Weinstein, and D. Slaney, Imported cases of Ross River virus disease in New Zealand ‐ a travel medicine perspective. Travel Med Infect Dis, 2012. 10(3): p. 129‐34. 163. Klapsing, P., et al., Ross River virus disease reemergence, Fiji, 2003‐2004. Emerg Infect Dis, 2005. 11(4): p. 613‐5. 164. Notifiable Conditions Reports. Queensland Department of Health. [Accessed September 23, 2018]; Available from: https://www.health.qld.gov.au/clinical‐ practice/guidelines‐procedures/diseases‐ infection/surveillance/reports/notifiable/annual/default.asp. 165. Age, T., 'Extraordinary numbers' of mosquitoes give way to record Ross River virus season in NSW, H. Alexander, Editor. 2017.
Chapter 8: References 220
166. News, A., Ross River virus warning follows increased numbers of mosquitoes in Victoria. 2017. 167. News, A., Ross River virus detected in Melbourne, possibly for first time. 2017. 168. News, A., Ross River virus cases spike in Western Australia after February floods. 2017. 169. Tong, S., et al., Climate variability, social and environmental factors, and ross river virus transmission: research development and future research needs. Environ Health Perspect, 2008. 116(12): p. 1591‐7. 170. Glass, K., Ecological mechanisms that promote arbovirus survival: a mathematical model of Ross River virus transmission. Trans R Soc Trop Med Hyg, 2005. 99(4): p. 252‐60. 171. Carver, S., et al., Influence of hosts on the ecology of arboviral transmission: potential mechanisms influencing dengue, Murray Valley encephalitis, and Ross River virus in Australia. Vector Borne Zoonotic Dis, 2009. 9(1): p. 51‐64. 172. Lindsay, M.D., et al., Ross River virus isolations from mosquitoes in arid regions of Western Australia: implication of vertical transmission as a means of persistence of the virus. Am J Trop Med Hyg, 1993. 49(6): p. 686‐96. 173. Kay, B.H., et al., Alphavirus infection in mosquitoes at the Ross River reservoir, north Queensland, 1990‐1993. J Am Mosq Control Assoc, 1996. 12(3 Pt 1): p. 421‐8. 174. Dhileepan, K., J.K. Azuolas, and C.A. Gibson, Evidence of vertical transmission of Ross River and Sindbis viruses (Togaviridae: Alphavirus) by mosquitoes (Diptera: Culicidae) in southeastern Australia. J Med Entomol, 1996. 33(1): p. 180‐2. 175. Frances, S.P., et al., Occurrence of Ross River virus and Barmah Forest virus in mosquitoes at Shoalwater Bay military training area, Queensland, Australia. J Med Entomol, 2004. 41(1): p. 115‐20. 176. Azuolas, J.K., et al., Isolation of Ross River virus from mosquitoes and from horses with signs of musculo‐skeletal disease. Aust Vet J, 2003. 81(6): p. 344‐7. 177. Ritchie, S.A., et al., Ross River virus in mosquitoes (Diptera:Culicidae) during the 1994 epidemic around Brisbane, Australia. J Med Entomol, 1997. 34(2): p. 156‐9. 178. Jardine, A., et al., Ross River virus risk associated with dispersal of Aedes (Ochlerotatus) camptorhynchus (Thomson) from breeding habitat into surrounding residential areas: muddy lakes, Western Australia. Am J Trop Med Hyg, 2014. 91(1): p. 101‐8. 179. Ryan, P.A., et al., Ross River virus disease clusters and spatial relationship with mosquito biting exposure in Redland Shire, southern Queensland, Australia. J Med Entomol, 2006. 43(5): p. 1042‐59. 180. Ryan, P.A., K.A. Do, and B.H. Kay, Definition of Ross River virus vectors at Maroochy Shire, Australia. J Med Entomol, 2000. 37(1): p. 146‐52. 181. Russell, R.C., et al., Mosquito (Diptera: Culicidae) and arbovirus activity on the south coast of New South Wales, Australia, in 1985‐1988. J Med Entomol, 1991. 28(6): p. 796‐804. 182. Jardine, A., et al., Ross River virus risk associated with dispersal of Aedes (Ochlerotatus) camptorhynchus (Thomson) from breeding habitat into
Chapter 8: References 221
surrounding residential areas: muddy lakes, Western Australia. Am J Trop Med Hyg, 2014. 91(1): p. 101‐108. 183. Ballard, J.W. and I.D. Marshall, An investigation of the potential of Aedes camptorhynchus (Thom.) as a vector of Ross River virus. Aust J Exp Biol Med Sci, 1986. 64 ( Pt 2): p. 197‐200. 184. Kay, B.H., Three modes of transmission of Ross River virus by Aedes vigilax (Skuse). Aust J Exp Biol Med Sci, 1982. 60(3): p. 339‐44. 185. Jeffery, J.A., B.H. Kay, and P.A. Ryan, Role of Verrallina funerea (Diptera: Culicidae) in transmission of Barmah Forest virus and Ross River virus in coastal areas of eastern Australia. J Med Entomol, 2006. 43(6): p. 1239‐47. 186. Jansen, C.C., et al., Arboviruses isolated from mosquitoes collected from urban and peri‐urban areas of eastern Australia. J Am Mosq Control Assoc, 2009. 25(3): p. 272‐8. 187. Hu, W., et al., Difference in mosquito species (Diptera: Culicidae) and the transmission of Ross River virus between coastline and inland areas in Brisbane, Australia. Environ Entomol, 2010. 39(1): p. 88‐97. 188. Jacups, S.P., et al., Predictive indicators for Ross River virus infection in the Darwin area of tropical northern Australia, using long‐term mosquito trapping data. Trop Med Int Health, 2008. 13(7): p. 943‐52. 189. Ryan, P.A., K.A. Do, and B.H. Kay, Spatial and temporal analysis of Ross River virus disease patterns at Maroochy Shire, Australia: association between human morbidity and mosquito (Diptera: Culicidae) abundance. J Med Entomol, 1999. 36(4): p. 515‐21. 190. Hu, W., et al., Mosquito species (Diptera: Culicidae) and the transmission of Ross River virus in Brisbane, Australia. J Med Entomol, 2006. 43(2): p. 375‐81. 191. Stephenson, E.B., et al., Interpreting mosquito feeding patterns in Australia through an ecological lens: an analysis of blood meal studies. Parasit Vectors, 2019. 12(1): p. 156. 192. Jansen, C.C., et al., Blood sources of mosquitoes collected from urban and peri‐ urban environments in eastern Australia with species‐specific molecular analysis of avian blood meals. Am J Trop Med Hyg, 2009. 81(5): p. 849‐57. 193. Boyd, A.M. and B.H. Kay, Assessment of the potential of dogs and cats as urban reservoirs of Ross River and Barmah Forest viruses. Aust Vet J, 2002. 80(1‐2): p. 83‐6. 194. Ryan, P.A., et al., Investigation of gray‐headed flying foxes (Pteropus poliocephalus) (Megachiroptera: Pteropodidae) and mosquitoes in the ecology of Ross River virus in Australia. Am J Trop Med Hyg, 1997. 57(4): p. 476‐82. 195. Carver, S., V. Sakalidis, and P. Weinstein, House mouse abundance and Ross River virus notifications in Victoria, Australia. Int J Infect Dis, 2008. 12(5): p. 528‐33. 196. Spradbrow, P.B., Letter: Experimental infection of sheep and pigs with Ross River virus. Aust Vet J, 1973. 49(8): p. 403‐4. 197. Whitehead, R.H., Experimental infection of vertebrates with Ross River and Sindbis viruses, two group A arboviruses isolated in Australia. Aust J Exp Biol Med Sci, 1969. 47(1): p. 11‐5.
Chapter 8: References 222
198. Koolhof, I.S. and S. Carver, Epidemic host community contribution to mosquito‐borne disease transmission: Ross River virus. Epidemiol Infect, 2017. 145(4): p. 656‐666. 199. Tong, S., Ross River virus disease in Australia: epidemiology, socioecology and public health response. Intern Med J, 2004. 34(1‐2): p. 58‐60. 200. Gatton, M.L., et al., Spatial‐temporal analysis of Ross River virus disease patterns in Queensland, Australia. Am J Trop Med Hyg, 2004. 71(5): p. 629‐35. 201. Hu, W., et al., Exploratory spatial analysis of social and environmental factors associated with the incidence of Ross River virus in Brisbane, Australia. Am J Trop Med Hyg, 2007. 76(5): p. 814‐9. 202. Tall, J.A., M.L. Gatton, and S. Tong, Ross River Virus Disease Activity Associated With Naturally Occurring Nontidal Flood Events in Australia: A Systematic Review. J Med Entomol, 2014. 51(6): p. 1097‐108. 203. Maelzer, D., et al., El Nino and arboviral disease prediction. Environ Health Perspect, 1999. 107(10): p. 817‐8. 204. Muhar, A., et al., The spatial distribution of ross river virus infections in Brisbane: Significance of residential location and relationships with vegetation types. Environ Health Prev Med, 2000. 4(4): p. 184‐9. 205. Kelly‐Hope, L.A., D.M. Purdie, and B.H. Kay, Differences in climatic factors between Ross River virus disease outbreak and nonoutbreak years. J Med Entomol, 2004. 41(6): p. 1116‐22. 206. Yu, W., et al., Projecting the impact of climate change on the transmission of Ross River virus: methodological challenges and research needs. Epidemiol Infect, 2014. 142(10): p. 2013‐23. 207. Hu, W., et al., Development of a predictive model for ross river virus disease in Brisbane, Australia. Am J Trop Med Hyg, 2004. 71(2): p. 129‐37. 208. Jardine, A., et al., Is there an association between dryland salinity and Ross River virus disease in southwestern Australia? Ecohealth, 2008. 5(1): p. 58‐68. 209. Ng, V., et al., Analysis and prediction of Ross River virus transmission in New South Wales, Australia. Vector Borne Zoonotic Dis, 2014. 14(6): p. 422‐38. 210. Tong, S. and W. Hu, Different responses of Ross River virus to climate variability between coastline and inland cities in Queensland, Australia. Occup Environ Med, 2002. 59(11): p. 739‐44. 211. Jacups, S.P., et al., Determining meteorological drivers of salt marsh mosquito peaks in tropical northern Australia. J Vector Ecol, 2015. 40(2): p. 277‐81. 212. Hu, W., et al., Bayesian spatiotemporal analysis of socio‐ecologic drivers of Ross River virus transmission in Queensland, Australia. Am J Trop Med Hyg, 2010. 83(3): p. 722‐8. 213. Gatton, M.L., B.H. Kay, and P.A. Ryan, Environmental predictors of Ross River virus disease outbreaks in Queensland, Australia. Am J Trop Med Hyg, 2005. 72(6): p. 792‐9. 214. Choi, Y.H., et al., Modelling the transmission dynamics of Ross River virus in Southwestern Australia. IMA J Math Appl Med Biol, 2002. 19(1): p. 61‐74. 215. Jardine, A., P.J. Neville, and M.D. Lindsay, Proximity to mosquito breeding habitat and Ross River virus risk in the Peel region of Western Australia. Vector Borne Zoonotic Dis, 2015. 15(2): p. 141‐6.
Chapter 8: References 223
216. Tong, S., et al., Climatic, high tide and vector variables and the transmission of Ross River virus. Intern Med J, 2005. 35(11): p. 677‐80. 217. Williams, C.R., S.R. Fricker, and M.J. Kokkinn, Environmental and entomological factors determining Ross River virus activity in the River Murray Valley of South Australia. Aust N Z J Public Health, 2009. 33(3): p. 284‐8. 218. Ritchie, E.G. and E.E. Bolitho, Australia's savanna herbivores: bioclimatic distributions and an assessment of the potential impact of regional climate change. Physiol Biochem Zool, 2008. 81(6): p. 880‐90. 219. Martin, G.A., et al., Climatic suitability influences species specific abundance patterns of Australian flying foxes and risk of Hendra virus spillover. One Health, 2016. 2: p. 115‐121. 220. Woodruff, R.E., et al., Early warning of Ross River virus epidemics: combining surveillance data on climate and mosquitoes. Epidemiology, 2006. 17(5): p. 569‐75. 221. Flasche, S., et al., The Long‐Term Safety, Public Health Impact, and Cost‐ Effectiveness of Routine Vaccination with a Recombinant, Live‐Attenuated Dengue Vaccine (Dengvaxia): A Model Comparison Study. PLoS Med, 2016. 13(11): p. e1002181. 222. Ritchie, S.A. and B.J. Johnson, Advances in Vector Control Science: Rear‐and‐ Release Strategies Show Promise... but Don't Forget the Basics. J Infect Dis, 2017. 215(suppl_2): p. S103‐S108. 223. Arunachalam, N., et al., Eco‐bio‐social determinants of dengue vector breeding: a multicountry study in urban and periurban Asia. Bull World Health Organ, 2010. 88(3): p. 173‐84. 224. Quintero, J., et al., Ecological, biological and social dimensions of dengue vector breeding in five urban settings of Latin America: a multi‐country study. BMC Infect Dis, 2014. 14: p. 38. 225. Bowman, L.R., S. Donegan, and P.J. McCall, Is Dengue Vector Control Deficient in Effectiveness or Evidence?: Systematic Review and Meta‐analysis. PLoS Negl Trop Dis, 2016. 10(3): p. e0004551. 226. Achee, N.L., et al., A critical assessment of vector control for dengue prevention. PLoS Negl Trop Dis, 2015. 9(5): p. e0003655. 227. Achee, N.L., et al., Alternative strategies for mosquito‐borne arbovirus control. PLoS Negl Trop Dis, 2019. 13(1): p. e0006822. 228. Manrique‐Saide, P., et al., Use of insecticide‐treated house screens to reduce infestations of dengue virus vectors, Mexico. Emerg Infect Dis, 2015. 21(2): p. 308‐11. 229. Choi, D.B., et al., Effect of Spatial Repellent Exposure on Dengue Vector Attraction to Oviposition Sites. PLoS Negl Trop Dis, 2016. 10(7): p. e0004850. 230. Lee, J., et al., Effect of the Topical Repellent para‐Menthane‐3,8‐diol on Blood Feeding Behavior and Fecundity of the Dengue Virus Vector Aedes aegypti. Insects, 2018. 9(2). 231. Wahid, I., et al., Integrated vector management with additional pre‐ transmission season thermal fogging is associated with a reduction in dengue incidence in Makassar, Indonesia: Results of an 8‐year observational study. PLoS Negl Trop Dis, 2019. 13(8): p. e0007606.
Chapter 8: References 224
232. Roiz, D., et al., Integrated Aedes management for the control of Aedes‐borne diseases. PLoS Negl Trop Dis, 2018. 12(12): p. e0006845. 233. WHO, Handbook for Integrated Vector Management. 2012: Geneva. 234. George, L., et al., Community‐Effectiveness of Temephos for Dengue Vector Control: A Systematic Literature Review. PLoS Negl Trop Dis, 2015. 9(9): p. e0004006. 235. Han, W.W., et al., Efficacy and community effectiveness of larvivorous fish for dengue vector control. Trop Med Int Health, 2015. 20(9): p. 1239‐1256. 236. Lazaro, A., et al., Community effectiveness of copepods for dengue vector control: systematic review. Trop Med Int Health, 2015. 20(6): p. 685‐706. 237. Marcos‐Marcos, J., et al., Impact, economic evaluation, and sustainability of integrated vector management in urban settings to prevent vector‐borne diseases: a scoping review. Infect Dis Poverty, 2018. 7(1): p. 83. 238. Nguyen‐Tien, T., A. Probandari, and R.A. Ahmad, Barriers to Engaging Communities in a Dengue Vector Control Program: An Implementation Research in an Urban Area in Hanoi City, Vietnam. Am J Trop Med Hyg, 2019. 100(4): p. 964‐973. 239. Tomerini, D.M., P.E. Dale, and N. Sipe, Does mosquito control have an effect on mosquito‐borne disease? The case of Ross River virus disease and mosquito management in Queensland, Australia. J Am Mosq Control Assoc, 2011. 27(1): p. 39‐44. 240. Aaskov, J.G., et al., Surveillance for Ross River virus infection using blood donors. American Journal of Tropical Medicine and Hygiene, 1998. 58(6): p. 726‐730. 241. Selden, S.M. and A.S. Cameron, Changing epidemiology of Ross River virus disease in South Australia. Med J Aust, 1996. 165(6): p. 313‐7. 242. Woodruff, R. and H. Bambrick, Climate change impacts on the burden of Ross River virus disease. Garnaut Climate Change Review. June 2008, Commonwealth of Australia: National Library of Australia. 243. Walker, L.J., et al., Mosquito and Virus Surveillance as a Predictor of Human Ross River Virus Infection in South‐West Western Australia: How Useful Is It? Am J Trop Med Hyg, 2018. 99(4): p. 1066‐1073. 244. van den Hurk, A.F., et al., Evolution of mosquito‐based arbovirus surveillance systems in Australia. J Biomed Biotechnol, 2012. 2012: p. 325659. 245. Foley, J.A., et al., Global consequences of land use. Science, 2005. 309(5734): p. 570‐4. 246. Steiger, D.B.M., S.A. Ritchie, and S.G.W. Laurance, Mosquito communities and disease risk influenced by land use change and seasonality in the Australian tropics. Parasites & Vectors, 2016. 9. 247. Fornace, K.M., et al., Local human movement patterns and land use impact exposure to zoonotic malaria in Malaysian Borneo. Elife, 2019. 8. 248. Kraemer, M.U.G., et al., Utilizing general human movement models to predict the spread of emerging infectious diseases in resource poor settings. Sci Rep, 2019. 9(1): p. 5151. 249. Zouache, K. and A.B. Failloux, Insect‐pathogen interactions: contribution of viral adaptation to the emergence of vector‐borne diseases, the example of chikungunya. Curr Opin Insect Sci, 2015. 10: p. 14‐21.
Chapter 8: References 225
250. Kraemer, M.U.G., et al., Past and future spread of the arbovirus vectors Aedes aegypti and Aedes albopictus. Nat Microbiol, 2019. 4(5): p. 854‐863. 251. Messina, J.P., et al., The current and future global distribution and population at risk of dengue. Nat Microbiol, 2019. 4(9): p. 1508‐1515. 252. Simoes, M., et al., Evaluation of accuracy and reliability of the plaque reduction neutralization test (micro‐PRNT) in detection of yellow fever virus antibodies. Biologicals, 2012. 40(6): p. 399‐404. 253. Borges, M.B., et al., Accuracy and repeatability of a micro plaque reduction neutralization test for vaccinia antibodies. Biologicals, 2008. 36(2): p. 105‐10. 254. Dorigatti, I., et al., Using Wolbachia for Dengue Control: Insights from Modelling. Trends Parasitol, 2018. 34(2): p. 102‐113. 255. Kamtchum‐Tatuene, J., et al., The potential role of Wolbachia in controlling the transmission of emerging human arboviral infections. Curr Opin Infect Dis, 2017. 30(1): p. 108‐116. 256. Buchman, A., et al., Broad dengue neutralization in mosquitoes expressing an engineered antibody. PLoS Pathog, 2020. 16(1): p. e1008103. 257. Harris, A.F., et al., Successful suppression of a field mosquito population by sustained release of engineered male mosquitoes. Nat Biotechnol, 2012. 30(9): p. 828‐30. 258. Wilder‐Smith, A., et al., Novel tools for the surveillance and control of dengue: findings by the DengueTools research consortium. Glob Health Action, 2018. 11(1): p. 1549930. 259. Kittayapong, P., et al., Application of eco‐friendly tools and eco‐bio‐social strategies to control dengue vectors in urban and peri‐urban settings in Thailand. Pathog Glob Health, 2012. 106(8): p. 446‐54. 260. Weaver, S.C., Prediction and prevention of urban arbovirus epidemics: A challenge for the global virology community. Antiviral Res, 2018. 156: p. 80‐ 84. 261. World Health Organization. Global Strategy for dengue prevention and control, 2012‐2020. August 2012: Geneva, Switzerland. 262. Boyd, A.M. and B.H. Kay, Vector competence of Aedes aegypti, Culex sitiens, Culex annulirostris, and Culex quinquefasciatus (Diptera: Culicidae) for Barmah Forest virus. J Med Entomol, 2000. 37(5): p. 660‐3. 263. Russell, R.C. and M.J. Geary, The susceptibility of the mosquitoes Aedes notoscriptus and Culex annulirostris to infection with dog heartworm Dirofilaria immitis and their vector efficiency. Med Vet Entomol, 1992. 6(2): p. 154‐8. 264. Kay, B.H., et al., Quantitative studies of the vector competence of Aedes aegypti, Culex annulirostris and other mosquitoes (Diptera: Culicidae) with Murray Valley encephalitis and other Queensland arboviruses. J Med Entomol, 1979. 16(1): p. 59‐66. 265. Jansen, C.C., S.A. Ritchie, and A.F. van den Hurk, The role of Australian mosquito species in the transmission of endemic and exotic West Nile virus strains. Int J Environ Res Public Health, 2013. 10(8): p. 3735‐52.
Chapter 8: References 226
Chapter 9: Appendices
Appendix A Published article: Stephenson EB, Murphy AK, Jansen CC, Peel AJ, McCallum, H. Interpreting mosquito feeding patterns in Australia through an ecological lens: an analysis of blood meal studies. Parasites and Vectors, 2019. DOI: 10.1186/s13071‐019‐3405‐z
Appendix B RRV field survey protocol
Appendix C Manuscript: Stephenson EB, Murphy AK, Jansen CC, Shivas M, McCallum, H, Peel AJ. Associations between Ross River virus disease and vector and vertebrate community ecology in Brisbane, Australia. Article published in Vector‐Borne and Zoonotic Diseases, 2020. DOI: 10.1089/vbz.2019.2585
Chapter 9: Appendices 227
Appendix A: Interpreting mosquito feeding patterns in Australia through an ecological lens: an analysis of blood meal studies
This article was produced in preparation for the field study conducted during this PhD. It reviews existing blood meal studies performed in Australia to date, provides an assessment of their contribution to understanding vector‐borne disease risks, and identifies research gaps and future directions.
See the online article here: https://parasitesandvectors.biomedcentral.com/articles/10.1186/s13071‐019‐ 3405‐z
Chapter 9: Appendices 228 Stephenson et al. Parasites Vectors (2019) 12:156 https://doi.org/10.1186/s13071-019-3405-z Parasites & Vectors
RESEARCH Open Access Interpreting mosquito feeding patterns in Australia through an ecological lens: an analysis of blood meal studies Eloise B. Stephenson1* , Amanda K. Murphy2, Cassie C. Jansen3, Alison J. Peel1 and Hamish McCallum1
Abstract Background: Mosquito-borne pathogens contribute signifcantly to the global burden of disease, infecting millions of people each year. Mosquito feeding is critical to the transmission dynamics of pathogens, and thus it is important to understanding and interpreting mosquito feeding patterns. In this paper we explore mosquito feeding patterns and their implications for disease ecology through a meta-analysis of published blood meal results collected across Australia from more than 12,000 blood meals from 22 species. To assess mosquito-vertebrate associations and identify mosquitoes on a spectrum of generalist or specialist feeders, we analysed blood meal data in two ways; frst using a novel odds ratio analysis, and secondly by calculating Shannon’s diversity scores. Results: We fnd that each mosquito species had a unique feeding association with diferent vertebrates, suggesting species-specifc feeding patterns. Broadly, mosquito species could be grouped broadly into those that were primar- ily ornithophilic and those that fed more often on livestock. Aggregated feeding patterns observed across Australia were not explained by intrinsic variables such as mosquito genetics or larval habitats. We discuss the implications for disease transmission by vector mosquito species classifed as generalist-feeders (such as Aedes vigilax and Culex annu- lirostris), or specialists (such as Aedes aegypti) in light of potential infuences on mosquito host choice. Conclusions: Overall, we fnd that whilst existing blood meal studies in Australia are useful for investigating mos- quito feeding patterns, standardisation of blood meal study methodologies and analyses, including the incorporation of vertebrate surveys, would improve predictions of the impact of vector-host interactions on disease ecology. Our analysis can also be used as a framework to explore mosquito-vertebrate associations, in which host availability data is unavailable, in other global systems. Keywords: Blood meal, Associations, Vector, Vertebrate, Disease risk, Blood-feeding
Background to take a blood meal from a source host and then to sub- Mosquitoes are the most important disease vector glob- sequently feed on a recipient host. Understanding the ally, responsible for infecting millions of people and feeding patterns of mosquitoes can inform disease man- animals annually with pathogens that infuence human agement strategies (such as targeted vector control to health, livestock and economic trade and wildlife bio- reduce vector-host contact) and can contribute to mod- diversity [1]. Mosquitoes comprise a broad taxonomic els forecasting future disease risk in human and animal group with more than 3000 species recognised across 40 populations [3]. genera [2], but not all species are involved in pathogen Mosquito host choice is complex; both intrinsic and transmission. Pathogen transmission requires a mosquito extrinsic factors can infuence feeding preference [3, 4]. Intrinsic variables can include genetics, whereby individ- *Correspondence: [email protected] uals are more likely to feed on the same host as previous 1 Environmental Futures Research Institute, Grifth University, Brisbane, generations [5, 6], and the nutritional state of the mos- QLD 4111, Australia quito, with nutrition-poor individuals being more likely Full list of author information is available at the end of the article
© The Author(s) 2019. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Stephenson et al. Parasites Vectors (2019) 12:156 Page 2 of 11
to feed on non-preferred hosts [4]. Extrinsic host-seeking this review if they were original peer-reviewed research behaviour is predominantly guided by detection of heat articles, undertaken on mainland Australia (e.g. [17] was and carbon dioxide (CO2), and is also afected by host undertaken in Badu Island in the Torres Strait and thus abundance, biomass, various odorants and chemicals excluded), analysed feld-collected mosquitoes that had that are released by hosts, and host defensive behaviour fed under natural conditions on free-living vertebrates [7–13]. Other extrinsic factors may include climatic vari- (e.g. [18] used tethered animal baits and was excluded), ables such as relative humidity, along with habitat charac- and assessed at least 3 potential vertebrate species (e.g. teristics that determine availability and diversity of hosts [19] only tested for a single fying fox species, and was [12, 14]. In addition to these broad intrinsic and extrinsic excluded). variables, evidence suggests that mosquitoes may adjust Te following information was extracted from identi- their host-seeking behaviours based on positive and neg- fed articles: the geographical area in which the study ative experiences, in essence, adapting feeding choices took place, including site location, bioregion of each site according to their individual circumstances [4]. (as defned by Tackway & Cresswell [20]), the mosquito Mosquito-host relationships in Australia are largely collection method used (including the year, month and understudied. Te island biogeography of Australia collection method/trap type), and the methods used to and its varied climatic zones and bioregions promote a determine the vertebrate origin of blood meals (includ- unique endemic biodiversity for mosquito and vertebrate ing the vertebrate species investigated, the source of host species. Along with a high diversity of native marsu- vertebrate reference samples and laboratory technique). pials, placental mammals and birds in Australia, there are Additional notes were made on stated limitations (if any) more than 300 species of mosquitoes described, many of of each paper and whether data on vertebrate abundance which are unique to the continent [15]. Te interactions and diversity was included. A database of blood meal between these populations of mosquitoes and vertebrate results was populated and is reported in Additional fle 1: hosts across diferent climatic zones provide opportuni- Table S1. ties for maintenance and emergence of mosquito-borne pathogens. Te transmission of numerous medically- Data analysis important arboviruses has been documented in Australia Mosquito‑vertebrate associations to date, including dengue (DENV), Ross River (RRV), Odds ratios were used to calculate the direction (positive Murray valley encephalitis (MVEV), Barmah Forest or negative) and strength of associations in the database (BFV) and Kunjin [16] viruses. between each mosquito species and vertebrate taxon. For Taking into account the complexity of contributing fac- this analysis, the blood meal origin data compiled from tors, critical analysis of the feeding patterns of mosqui- the literature were aggregated such that each vertebrate toes may represent an important approach to explore species was grouped into the broader taxonomic groups disease risks for both human and animal populations. of humans, Carnivora (cats, dogs and foxes), Aves (all Tis study aims to synthesise existing literature describ- birds), Diprotodontia (all possum and kangaroo species), ing blood meal studies in Australia, specifcally assess- Artiodactyla (cows, sheep, pigs and goats) and Perisso- ing the most likely mosquito-host associations, and the dactyla (horses). Flying fox [21], rodent [22, 23] and rab- diversity of feeding patterns for common mosquito spe- bit [22, 24] species were excluded from this analysis, as cies. In light of these feeding patterns, we discuss broad the sample size of blood meals from these species was too implications for disease ecology. small (either one or two studies, with these species com- prising less than 9% of blood-meal origins within each). Methods To be included in the analysis, mosquito species needed Data collection to meet all minimum data criteria of (i) their blood meal Original research articles were systematically searched origins being reported more than twice in the literature; by using the following search terms in diferent combi- and (ii) having an arbitrary minimum of 35 blood meals nation across fve search engines (Web of Science, Pro- identifed. Quest, Science Direct, PubMed and Google Scholar): Log odds ratios were calculated between each mos- ‘bloodmeal*’, ‘blood meal’, blood-meal’, ‘feeding’, ‘habit’, quito and vertebrate taxon using two by two feeding ‘pattern*’, ‘preference*’, ‘interaction*’, ‘mosquito*’, ‘vec- frequency tables, derived from the raw data (Additional tor*’, ‘vector-host’, ‘host*’, ‘vertebrate*’, ‘animal*’ and ‘Aus- fle 1: Table S2). Positive log odds ratios indicate a posi- tralia’. Te asterisk (*) operator was used as a wildcard to tive feeding association between a given mosquito species search for all possible variations of keywords. We then and vertebrate host, whereby there is a higher likelihood manually searched the reference lists of papers to iden- than random chance that a blood meal of that mosquito tify additional relevant articles. Papers were included in species would originate from the given vertebrate taxon. Stephenson et al. Parasites Vectors (2019) 12:156 Page 3 of 11
Te greater the log odds ratio, the stronger the feeding To collect blood-fed mosquitoes, most studies (n = association. Conversely, a negative log odds ratio suggests 8) used Centers for Disease Control (CDC) CO2-baited a negative feeding association, whereby there is a lower miniature light traps [28], supplemented with 1-octen- likelihood that a blood meal from the mosquito species 3-ol in some cases (Table 1). Other methods included would originate from the given vertebrate taxon. Log unbaited BioGent® (BG) sentinel traps [22] and aspira- odds ratios close to 0 indicate no association between the tion of resting sites [21, 29, 30]. One study [29] also used mosquito species and vertebrate taxon. vehicle-mounted traps. To analyse blood meals, early Te log odds ratios were plotted in a heatmap chart studies employed precipitin tests [29–31] and serological and sorted using hierarchical clustering. Te cluster- gel difusion techniques [32, 33] (Table 1). More recent ing grouped mosquitoes with similar feeding patterns studies adopted enzyme-linked immunosorbent assay together by similarity in log odds ratio across all verte- (ELISA) and various molecular techniques including pol- brate taxa. All calculations and graphs were generated ymerase chain reaction (PCR) and gene sequencing [22– using R software, with packages gplots and RColorBrewer 24, 34]. Vertebrate reference sources most commonly [25] with modifed script from Raschka [26]. employed in immunoassays and gel difusion techniques were commercially-available anti-sera and included Mosquito feeding diversity horse, rabbit, rat, dog, chicken, cat, bird, kangaroo, cow We used the Shannonʼs diversity index to place mosquito and pig. Molecular studies which included wildlife used species on a spectrum between generalist or specialist vertebrate references provided through wildlife hospi- feeders. Te inclusion criteria for this analysis were that tals, zoos and roadkill. Tese studies also included DNA each mosquito species needed to have fed on greater sequence data available on GenBank. than three vertebrate species and had to have a minimum number of 10 blood meals analysed. Vertebrates were Mosquito‑vertebrate associations not aggregated by taxonomic group in this analysis but Of the 10 studies of blood meal origins, data on 41 mos- remained at the level reported in the literature (mostly as quito species were reported, and data on 12 of these met species but, for the case of birds, several studies reported the criteria to be included in analysis: Aedes norman- as class). A total of 15 vertebrate species were included ensis, Ae. notoscriptus, Ae. procax, Ae. vigilax, Anoph- in this analysis as blood-meal origins, and 13,934 blood eles annulipies, An. bancroftii, Coquillettidia linealis, meals from 21/41 mosquito species met the criteria Cq. xanthogaster, Culex annulirostris, Cx. sitiens, Cx. (Additional fle 1: Table S1). quinquefasciatus and Mansonia uniformis (Fig. 2). All Shannonʼs diversity index was calculated for each mos- species, except Ae. procax, showed signifcant positive quito species [27] and expressed as an h-index. A higher associations with at least one vertebrate host. Te strong- h-index is associated with a greater feeding diversity, as est positive log odds ratio was between Cx. annulirostris it suggests a mosquito species has fed on a greater num- and the Diprotodontia taxa (possums and kangaroos; log ber of vertebrate species and/or feeds evenly across ver- odds ratio (LOR) = 2.77), followed by Ma. uniformis and tebrates. Conversely, a lower h-index suggests mosquito humans (LOR = 2.2). All mosquito species except Ae. species have a low feeding diversity, and are associated vigilax and Cq. xanthogaster had strong negative asso- with feeding on fewer vertebrate species and/or a greater ciation with at least one vertebrate taxon. Te strongest number of feeds on a small number of vertebrates. negative log odds ratio was between Cx. quinquefasciatus Within this dataset, we categorised an h-index in the top and Diprotodontia (LOR = -3.8), followed by Ae. noto- quartile as ‘high feeding diversity’, whilst an h-index in scriptus and Artiodactyla (cows, sheep, pigs and goats; the lowest quartile was considered a ‘low feeding diver- LOR = -2.8). sity’. Shannon’s diversity index was calculated in Excel. Te mosquito species clustered together in two broad groups. Te frst cluster group consisted of seven mos- Results quito species (Cx. quinquefasciatus, Ae. vigilax, Cq. xan- Characteristics of the selected studies thogaster, Ae. procax, Cq. linealis, Cx. sitiens and Ae. We identifed ten papers that met the search criteria, notoscriptus), of which most shared a negative associa- comprising 14,044 mosquito blood meals across 48 mos- tion with the Artiodactyla (6 of the 7 species) and Dipro- quito species. Study characteristics and methodologies todontia vertebrates (5/7), and a positive association with are summarised in Table 1. Tese studies took place at humans (7/7) and Aves (6/7). Within this cluster, Ae. 32 sites across 14 bioregions, in all mainland states and vigilax and Cq. xanthogaster were in the same clade and territories in Australia (Fig. 1). Te selected studies were shared a strong positive association with Perissodactyla. undertaken over a 62 year-period, from 1954 to 2016. Coquillettidia linealis and Cx. sitiens also shared a clade Stephenson et al. Parasites Vectors (2019) 12:156 Page 4 of 11 cat, bird, kangaroo, kangaroo, cat, bird, pig cow, dog, chicken, cat, bird, chicken, cat, bird, dog, 50 pig, cow, kangaroo, species wild bird rabbit, horse, donkey, donkey, rabbit, horse, dog, human, cat, fox, brushtail possum, quokka, Western grey mouse, kangaroo, chicken, duck pig, chicken, brushtail pig, and ringtail possum, koala, kangaroo, red lorikeet,rainbow galah, magpie Australian cow horse, dog, dog, chicken, cat, bird, chicken, cat, bird, dog, pig cow, kangaroo, horse, human, brushtail horse, possum, fying fox species marsupial, chicken marsupial, pig, dog, cat, man, dog, pig, bat, rodent, horse, cow, carnivore marsupial, Horse, rat, human, dog, rat, human, dog, Horse, Vertebrate species tested Vertebrate Horse, rabbit, rat, human, Horse, Cow, sheep, goat, pig, goat, pig, sheep, Cow, Dog, cat, cow, sheep, sheep, cat, cow, Dog, bird, Human, kangaroo, Horse, rabbit, rat, human, Horse, Bird, kangaroo, cat, dog, cat, dog, kangaroo, Bird, Cow, horse, dog, human, dog, horse, Cow, Reptile, amphibian, bird, amphibian, bird, Reptile, - - - using avian- and using avian- mammalian-specifc gene primers, sequencing using vertebrate, using vertebrate, and mamma - avian gene lian primers, sequencing assay assay assay Double-antibody ELISA Blood-meal analysis method Indirect ELISA; PCR Double-antibody ELISA Cytochrome b PCR Cytochrome Gel difusion immuno Gel difusion immuno Gel difusion immuno Precipitin test Precipitin Precipitin test Precipitin 199 (8) 763 (15) 865 (10) 1128 (1) 1180 (15) 2606 (29) 2582 (15) 1628 (18) 5431 (16) No. of blood meals No. of species) (no. analysed - 2 CO + / − octe + / − OCT + / − OCT 2 2 2 2 2 2 CO CO CO CO CO CO nol traps specifed) and vehicle specifed) and vehicle trap mounted sites EVS + Mosquito collection method EVS + Unbaited BG sentinel Unbaited CDC + CDC, EVS + CDC + CDC + Light trap (not EVS + Aspiration of resting of resting Aspiration - (1993–2004) 2015) (2000–2001) (2005–2008) ber (2002–2006) 2001) (1974–1976) (1998–2000) All year round round All year Collection months (years) August–May (2006– August–May September–April September–April All year round round All year February, March, Octo March, February, January–May (1995– All year round round All year All year round round All year February (1976) February Urban and rural Site habitat typeSite Urban and rural Urban Urban Rural Rural Rural Rural Rural Australia) City, Murray River val - City, ley (South Australia) Brisbane (Queensland), (NewNewcastle South (New Sydney Wales), South Wales) Territory) (Queensland) Territory) (Queensland) 32 sites (Western32 sites Site name (state) Site Adelaide Hills, Adelaide Hills, Adelaide Adelaide Brisbane (Queensland) Cairns (Queensland), Cairns (Queensland), (Queensland, Northern(Queensland, Gulf Plains York, Cape Beatrice Hill (Northern Shoalwater Bay Bay Shoalwater Charleville (Queensland) Location and methods for mosquito collection and blood meal analysis for the studies included in this review collection mosquito and blood meal analysis for and methods for Location 24 ] [ Johansen et al. 1 Table Reference al. [ 23 ] et al. Flies Kay [ 21 ] et al. al. [ 22 ] Jansen et al. al. [ 34 ] Hall-Mendelin et al. [ 33 ] van den Hurk et al. al. [ 29 ] Muller et al. al. [ 32 ] et al. Frances Kay [ 30 ] et al. Stephenson et al. Parasites Vectors (2019) 12:156 Page 5 of 11 - dog, rabbit, horse, mar rabbit, horse, dog, supial (unspecifed) Vertebrate species tested Vertebrate Human, chicken, cow, Human, chicken, cow, Blood-meal analysis method Precipitin test Precipitin 1231 (15) 14,044 (48) No. of blood meals No. of species) (no. analysed sites Mosquito collection method Aspiration of resting of resting Aspiration (1951–1952) Collection months (years) December–January December–January Site habitat typeSite Mostly rural , carbon dioxide; CDC, centers for disease control; BG, BioGent; PCR, polymerase chain reaction; ELISA, enzyme-linked BG, BioGent; PCR, polymerase disease control; for octane CDC, centers immunosorbent assay;, carbon dioxide; OCT, 2 CO South Wales), Texas Texas South Wales), Golburn(Queensland), (Victoria),Valley Canberra (Australian Territory) Capital Site name (state) Site Moree, Hornsby (New Moree, (continued) 1 Table survey;: EVS, vector Abbreviations encephalitis Reference al. [ 31 ] et al. Lee Total Stephenson et al. Parasites Vectors (2019) 12:156 Page 6 of 11
Fig. 1 Bioregions in which blood meal studies took place (indicated in red) across Australia (derived from Google Map Data ©) and strong negative association with both Artiodactyla was reported in fve mosquito species; of these, Ae. vigi- and Carnivora. lax had the highest diversity (h-index = 2.17). Te second major cluster group consisted of fve mos- quito species (Cx. annulirostris, Ae. normanesis, An. Discussion annulipies, An. bancroftii and Ma. uniformis). Tese spe- Our analysis revealed that each mosquito species had a cies all shared a positive association with Artiodactyla unique feeding association with diferent vertebrates, and a negative association with Aves. Culex annulirostris suggesting species-specifc feeding patterns. Te hierar- and Ae. normanesis were on the same clade and shared chical clustering from the odds ratio analysis sorted mos- a strong positive association with Perissodactyla and quitoes into two broad groups: mosquitoes that either Diprotodontia. Although they both also had a negative had a positive association with birds (Aves) and negative association with Aves, Carnivora and humans, this was association with livestock (Artiodactyla), or vice-versa. strongest only for Ae. normanensis. Anopheles annulipies Interpreting the feeding patterns of these particular mos- and Ma. uniformis were on a single clade and were both quito species is important, given that at least half of these had high associations with humans. mosquitoes have been found to be competent vectors for notifable arboviruses in Australia [35–40], whilst the Mosquito feeding diversity other half have been demonstrated to carry some viruses, Twenty-two mosquito species met the criteria for inclu- although their ability to transmit them has not been fully sion in the diversity analysis based on Shannonʼs index investigated [41–43]. (Fig. 3), comprising 12,424 individual blood meals in Intrinsic drivers of mosquito host choices (such as total. Te median h-index reported across all species genetics, larval ecology and dispersal) did not explain was 1.40, and the mean was 1.34. Low feeding diver- feeding patterns in this analysis. Specifcally, mosquito sity (h-index = < 0.99) was observed in fve mosquito species did not group together by taxonomic related- species; of which Ae. aegypti had the lowest diveristy ness (e.g. genus). Studies examining the efect of genet- (h-index = 0.72). High feeding diversity (h-index = > 1.64) ics on mosquito host choices have found that ofspring Stephenson et al. Parasites Vectors (2019) 12:156 Page 7 of 11
Fig. 2 Feeding associations between Australian mosquito species and vertebrate taxa. Log odds ratio for mosquito species (right hand side) indicate feeding likelihood on vertebrate taxa (bottom). Mosquito species are sorted in the chart using a hierarchal cluster (left) according to how similar their vertebrate feeding patterns are
Fig. 3 Shannon’s diversity (h-index) of blood meal origins for Australian mosquitoes, error bars represent the standard error for all measures Stephenson et al. Parasites Vectors (2019) 12:156 Page 8 of 11
are more likely to feed on the same host as previous despite being exposed to the same vertebrates in a single generations [5, 6]. However, this has only been demon- location. strated within-species and is unlikely to be important In the odds ratio analysis, Cx. annulirostris exhibited between species belonging to same genus, particularly an unexpected feeding pattern. Tis species is consid- since potential for rapid evolution (due to short genera- ered an important vector for medically-important arbo- tion span) likely reduces the infuence that taxonomic viruses [36, 49, 50]; however, the meta-analyses which relatedness may have on mosquito feeding host behav- included more than 5700 blood meals, found that Cx. iour. Another intrinsic factor, mosquito larval ecology, annulirostris had only a weak feeding association with may partially explain some clustering. For example, Ae. humans (LOR = -0.74). Tis is consistent with an early normanensis and Cx. annulirostris grouped together and feld study assessing mosquito feeding preferences using larvae of both typically inhabit inland freshwater; simi- live baits, in which Cx. annulirostris preferred cows, pigs larly, the larval habitat of both An. bancroftii and Ma. and dogs more than humans [37]. Te analysis based on uniformis is freshwater swamps. Although this pattern Shannonʼs diversity, along with other studies, have iden- did not explain all clusters, it implies that local environ- tifed Cx. annulirostris as a generalist feeder with plastic mental infuences, at least partially, drive mosquito host feeding patterns that may shift temporally or spatially choice. Tis is perhaps not surprising when considering [51, 52]. Tis knowledge, in combination with the wide- potential limitations on dispersal from larval habitats for spread distribution of Cx. annulirostris across Australia, various mosquito species. For example, Ae. vigilax is rec- suggests that localised studies of Cx annulirostris feeding ognised as having large dispersal capability, being found are required to assess the role the species plays in disease more than 50 km from potential saltwater larval habitats, transmission for which it is theoretically an important albeit likely wind-assisted in some cases [44]. Tis high vector. dispersal potential suggests that Ae. vigilax can move In addition to Cx. annulirostris, Aedes vigilax and Ae. readily between locations, allowing feeding on a diversity notoscriptus were identifed as generalists due to their of vertebrate taxa, as refected in our feeding diversity high diversity scores in the Shannon’s diversity analy- analysis. As such, whilst genetics and larval habitats may sis. International studies [53–55] suggest that generalist be important within species on a local scale, they do not feeders are capable of playing a role as bridge vectors due explain the aggregated feeding patterns observed across to their ability to acquire pathogens from animal hosts, Australia. and subsequently transmitting the pathogen to humans. Extrinsic variables, such as species abundance and Bridge vectors are particularly important for enzootic diversity, explain in part some of the feeding associa- amplifcation of arboviruses and are often associated with tions in this analysis, but not all. Mosquitoes have com- outbreaks [53]. For the species identifed in this analy- plex interactions with their environment. Tus, factors sis as generalists, they have been demonstrated to be broader than vertebrate abundance alone are important competent vectors of zoonotic arboviruses in Australia to consider for mosquito feeding patterns. For example, [45, 56–58], and as such should be closely monitored to mosquito fying/resting height has been linked to host reduce transmission between vectors and humans. feeding patterns [45–48]. In two Australian studies, more Te disease ecology associated with specialist feeders is Cx. sitiens and Cx. quiquefasciatus were caught in traps also important to consider. Here we identifed Ae. aegypti set at least 8 m of the ground, whilst a higher abundance as having the lowest feeding diversity, indicating the spe- of Ae. vigilax were found in traps 1.5 m of the ground, cies is a specialist feeder. Indeed, more than 70% of the for the same locations [47, 48]. In our meta-analysis, Cx. blood meals originated from humans. Te anthropophilic sitiens and Cx. quinquefasciatus had strong positive asso- feeding observed in Ae. aegypti is similar to that reported ciations with blood meals originating from tree dwelling in international studies, where 80–99% of all blood meals bird species (i.e. Australasian fgbirds Sphecotheres vieil- are human in origin [59, 60]. Tis feeding pattern for Ae. loti, common myna Sturnus tristis and helmeted friar- aegypti is consistent with its role as an important vector birds Philemon buceroides [22]), whilst Ae. vigilax had the of several arboviruses which are transmitted between strongest positive associations with ground-dwelling spe- humans without an animal reservoir, including dengue, cies (horses and humans). Tis could suggest that whilst Zika and chikungunya viruses. Interestingly, although the overall vertebrate abundance within a given environment importance of Ae. aegypti is recognised, under laboratory can infuence the availability of a particular host, mosqui- conditions the species has been observed to demonstrate toes are highly mobile and may seek a blood meal across relatively poor transmission rates for DENV, when com- ecological niches within a given habitat. Tus, diferent pared to other mosquito species [61, 62]. In this case, mosquito species can exhibit diferent feeding patterns being a specialist feeder, preferring mainly humans, is Stephenson et al. Parasites Vectors (2019) 12:156 Page 9 of 11
what determines the status of Ae. aegypti as an important likely infuenced by a suite of intrinsic and extrinsic vari- disease vector, rather than its competence [63]. ables. Broader ecological considerations alongside these feeding patterns could be useful for the interpretation of Future directions these complex biological systems, but at present data avail- An absence of data on host availability in the regions able to do this are limited. Future studies should utilise where mosquitoes were collected limits inferences on multidisciplinary approaches to collect data on vertebrate host preference specifcally. Of the blood meal studies communities in parallel with mosquito communities. More reviewed here, only one considered host abundance [21]. data from both top-down (broad assessments of blood Tat study assessed abundance through a local resident meals) and bottom-up approaches (specialised host choice survey on the number of pets, people and estimated experiments) are needed in conjunction with modelling number of possums in the vicinity, adding confdence to techniques to bring these data together for meaningful the interpretation of vector-feeding patterns. Such col- interpretation of arbovirus transmission risk in Australia. lection of host ecology data in conjunction with blood- fed mosquitoes can be considerably labour-intensive; Additional fle however, it provides a more thorough assessment of how host abundance and biomass may infuence observed mosquito feeding patterns and informs the selection of Additional fle 1: Table S1. Reported blood-meal results for Australian mosquito species. Numbers in the columns representing number of blood appropriate reference samples against which to com- meals for each vertebrate. Table S2. Derivation of 2 2 contingency pare blood meals in the laboratory. Although limited in tables for mosquito feeding preferences. × their application, other blood-meal studies [64, 65] have utilised databases, such as the Atlas of Living Australia Acknowledgements (ALA) or the Global Biodiversity Information Facility Not applicable.
(GBIF), to identify potential available vertebrates in the Funding absence of formal vertebrate surveys. Whilst they are no ES was supported by Australian Government Research Training Program substitute for vertebrate surveys, these datasets could be Scholarship. benefcial for noting the presence of common hosts in Availability of data and materials future blood meal studies but are limited in estimating The datasets used and/or analysed during the present study are available from host density or true absence of a given species. the corresponding author upon reasonable request.
Although a range of reference vertebrates were often Authors’ contributions included in Australian blood-meal studies, they were ES collected, analysed, interpreted papers and was a major contributor in the rarely a true representation of the vertebrates available to writing of the manuscript. AM, CC, AP and HM contributed to the interpreta- tion of results and writing of the manuscript. All authors read and approved mosquitoes for feeding. At present there are large gaps in the fnal manuscript. understanding the role of cryptic, migratory or smaller mammalian species in mosquito feeding patterns. For Ethics approval and consent to participate example, only two studies included rabbits [22, 24] and Not applicable. rodents [22, 23] in their analysis. Despite their small Consent for publication size, rabbits and rodents were identifed to be the origin Not applicable. of blood meals for Cx. sitiens and Cq. linealis [22]. Mos- Competing interests quito-rodent associations have also been identifed in the The authors declare that they have no competing interests. literature, where by at least 27% of mice were seropositive to RRV [58, 66]. It is therefore important that, despite Publisher’s Note small body size, rats and rodents are included in future Springer Nature remains neutral with regard to jurisdictional claims in pub- investigations of mosquito blood meals. lished maps and institutional afliations. Author details Conclusions 1 Environmental Futures Research Institute, Grifth University, Brisbane, QLD 4111, Australia. 2 QIMR, Herston, QLD 4006, Australia. 3 Communicable Improved understanding of mosquito feeding patterns Diseases Branch, Department of Health, Queensland Government, Herston, can lead to better management and risk predictions for QLD 4006, Australia. medically important arboviruses. Here we fnd that of the Received: 15 November 2018 Accepted: 20 March 2019 Australian mosquito species tested, each had a unique feeding pattern; however, the particular specialist or gen- eralist feeding patterns of mosquito species could be a key determinant of the risk they pose for human disease. Tese patterns, and the resulting human disease risk, are Stephenson et al. Parasites Vectors (2019) 12:156 Page 10 of 11
References 26. Raschka S. Instant heat maps in R: how-to. Birmingham UK: Packt 1. Gubler DJ. Human arbovirus infections worldwide. Ann N Y Acad Sci. Publishing Ltd.; 2013. 2001;951:13–24. 27. Morris EK, Caruso T, Buscot F, Fischer M, Hancock C, Maier TS, 2. Wongkamchai S, Khongtak P, Leemingsawat S, Komalamisra N, Junsong et al. Choosing and using diversity indices: insights for ecological N, Kulthanan K, et al. Comparative identifcation of protein profles and applications from the German Biodiversity Exploratories. Ecol Evol. major allergens of saliva, salivary gland and whole body extracts of mos- 2014;4:3514–24. quito species in Thailand. Asian Pac J Allergy Immunol. 2010;28:162–9. 28. Sudia W, Chamberlain R. Battery-operated light trap, an improved model. 3. Lyimo IN, Ferguson HM. Ecological and evolutionary determinants of Mosq News. 1962;22:126–9. host species choice in mosquito vectors. Trends Parasitol. 2009;25:189–96. 29. Muller M, Murray M, Edwards J. Blood-sucking midges and mosquitoes 4. Takken W, Verhulst NO. Host preferences of blood-feeding mosquitoes. feeding on mammals at Beatrice Hill, NT. Aust J Zool. 1981;29:573–88. Annu Rev Entomol. 2013;58:433–53. 30. Kay B, Boreham P, Williams G. Host preferences and feeding patterns of 5. Mukwaya L. Genetic control of feeding preferences in the mosquitoes mosquitoes (Diptera: Culicidae) at Kowanyama, Cape York Peninsula, Aedes (Stegomyia) simpsoni and aegypti. Physiol Entomol. 1977;2:133–45. northern Queensland. Bull Entomol Res. 1979;69:441–57. 6. Ulloa García A, Arredondo Jiménez JI, Fernández Salas I, Rodríguez MH, 31. Lee DJ, Clinton K, O’gower A. The blood sources of some Australian González Cerón L. Innate host selection in Anopheles vestitipennis from mosquitoes. Aust J Biol Sci. 1954;7:282–301. southern Mexico. J Am Mosq Control Assoc. 2004;20:337–41. 32. Frances SP, Cooper RD, Rowclife KL, Chen N, Cheng Q. Occurrence of 7. Edman JD, Taylor DJ. Culex nigripalpus: seasonal shift in the bird-mammal Ross River virus and Barmah Forest virus in mosquitoes at Shoalwa- feeding ratio in a mosquito vector of human encephalitis. Science. ter Bay military training area, Queensland, Australia. J Med Entomol. 1968;161:67–8. 2004;41:115–20. 8. Balenghien T, Fouque F, Sabatier P, Bicout DJ. Horse-, bird-, and human- 33. van den Hurk AF, Johansen CA, Zborowski P, Paru R, Foley PN, Beebe seeking behavior and seasonal abundance of mosquitoes in a West Nile NW, et al. Mosquito host-feeding patterns and implications for Japanese virus focus of southern France. J Med Entomol. 2006;43:936–46. encephalitis virus transmission in northern Australia and Papua New 9. Kilpatrick AM, Kramer LD, Jones MJ, Marra PP, Daszak P. West Nile virus Guinea. Med Vet Entomol. 2003;17:403–11. epidemics in North America are driven by shifts in mosquito feeding 34. Hall-Mendelin S, Jansen CC, Cheah WY, Montgomery BL, Hall RA, Ritchie behavior. PLoS Biol. 2006;4:e82. SA, et al. Culex annulirostris (Diptera: Culicidae) host feeding patterns and 10. Eiras AE, Jepson P. Responses of female Aedes aegypti (Diptera: Culicidae) Japanese encephalitis virus ecology in northern Australia. J Med Entomol. to host odours and convection currents using an olfactometer bioassay. 2012;49:371–7. Bull Entomol Res. 1994;84:207–11. 35. Watson TM, Kay BH. Vector competence of Aedes notoscriptus (Diptera: 11. Bowen M. The sensory physiology of host-seeking behavior in mosqui- Culicidae) for Barmah Forest virus and of this species and Aedes aegypti toes. Ann Rev Entomol. 1991;36:139–58. (Diptera: Culicidae) for dengue 1–4 viruses in Queensland. Australia. J 12. Clements A. The biology of mosquitoes. London: Chapman & Hall; 1992. Med Entomol. 1999;36:508–14. 13. Day JF, Ebert KM, Edman JD. Feeding patterns of mosquitoes (Diptera: 36. Boyd AM, Kay BH. Vector competence of Aedes aegypti, Culex sitiens, Culex Culicidae) simultaneously exposed to malarious and healthy mice, includ- annulirostris, and Culex quinquefasciatus (Diptera: Culiciclae) for Barmah ing a method for separating blood meals from conspecifc hosts. J Med Forest virus. J MedEntomol. 2000;37:660–3. Entomol. 1983;20:120–7. 37. Ryan PA, Kay BH. Vector competence of mosquitoes (Diptera: Culicidae) 14. Olanga EA, Okal MN, Mbadi PA, Kokwaro ED, Mukabana WR. Attraction of from Maroochy Shire, Australia, for Barmah Forest virus. J Med Entomol. Anopheles gambiae to odour baits augmented with heat and moisture. 1999;36:856–60. Malar J. 2010;9:6. 38. Wells PJ, Russell RC, Cloonan MJ. Investigating vector competence of 15. Webb C, Russell R, Doggett S. A guide to mosquitoes of Australia. New Culex annulirostris and Aedes vigilax for Ross River virus and other alpha- York: CSIRO Publishing; 2016. and bunyaviruses. Arbov Res Aust. 1993;6:10–4. 16. van den Hurk AF, Jansen CC. Arboviruses of Oceania. In: Loukas A, editor. 39. Kay BH, Fanning ID, Carley JG. Vector competence of Culex pipiens Neglected tropical diseases - Oceania. Berlin: Springer; 2016. p. 193–235. quinquefasciatus for Murray Valley encephalitis, Kunjin, and Ross River 17. van den Hurk AF, Nisbet DJ, Johansen CA, Foley PN, Ritchie SA, Mackenzie viruses from Australia. Am J Trop Med Hyg. 1982;31:844–8. JS. Japanese encephalitis on Badu Island, Australia: the frst isolation of 40. Kay BH, Carley JG, Fanning ID, Filippich C. Quantitative studies of the vec- Japanese encephalitis virus from Culex gelidus in the Australasian region tor competence of Aedes aegypti, Culex annulirostris and other mosquitos and the role of mosquito host-feeding patterns in virus transmission (Diptera, Culicidae) with Murray Valley encephalitis and other Queensland cycles. Trans R Soc Trop Med Hyg. 2001;95:595–600. arboviruses. J Med Entomol. 1979;16:59–66. 18. Muller M, Murray M. Blood-sucking fies feeding on sheep in eastern 41. Jefery JA, Ryan PA, Lyons SA, Kay BH. Vector competence of Coquillet- Australia. Aust J Zool. 1977;25:75–85. tidia linealis (Skuse) (Diptera: Culicidae) for Ross River and Barmah Forest 19. Ryan PA, Martin L, Mackenzie JS, Kay BH. Investigation of gray-headed viruses. Aust J Entomol. 2002;41:339–44. fying foxes (Pteropus poliocephalus) (Megachiroptera: Pteropodidae) and 42. van den Hurk AF, Nisbet DJ, Foley PN, Ritchie SA, Mackenzie JS, Beebe mosquitoes in the ecology of Ross River virus in Australia. Am J Trop Med NW. Isolation of arboviruses from mosquitoes (Diptera: Culicidae) col- Hyg. 1997;57:476–82. lected from the Gulf Plains region of Northwest Queensland, Australia. J 20. Thackway R, Cresswell I. A bioregional framework for planning Med Entomol. 2002;39:786–92. the national system of protected areas in Australia. Nat Areas J. 43. Harley D, Phillips D, Ritchie S, Atkin C, van den Hurk A. Isolation of Ross 1997;17:241–7. River virus from mosquitoes trapped in North Queensland. Arbov Res 21. Kay BH, Boyd AM, Ryan PA, Hall RA. Mosquito feeding patterns and Aust. 1997;7:92–4. natural infection of vertebrates with Ross River and Barmah Forest viruses 44. Chapman HF, Hughes JM, Jennings C, Kay BH, Ritchie SA. Population in Brisbane, Australia. Am J Trop Med Hyg. 2007;76:417–23. structure and dispersal of the saltmarsh mosquito Aedes vigilax in 22. Jansen CC, Webb CE, Graham GC, Craig SB, Zborowski P, Ritchie SA, et al. Queensland, Australia. Med Vet Entomol. 1999;13:423–30. Blood sources of mosquitoes collected from urban and peri-urban envi- 45. Lee HI, Seo BY, Shin EH, Burkett DA, Lee WJ, Shin YH. Study of fying height ronments in eastern Australia with species-specifc molecular analysis of of culicid species in the northern part of the Republic of Korea. J Am avian blood meals. Am J Trop Med Hyg. 2009;81:849–57. Mosq Cont Assoc. 2006;22:239–45. 23. Flies EJ, Flies AS, Fricker SR, Weinstein P, Williams CR. Regional compari- 46. Swanson D, Adler P. Vertical distribution of haematophagous Dip- son of mosquito bloodmeals in South Australia: implications for Ross tera in temperate forests of the southeastern USA. Med Vet Entomol. River virus ecology. J Med Entomol. 2016;53:902–10. 2010;24:182–8. 24. Johansen CA, Power SL, Broom AK. Determination of mosquito (Dip- 47. Jansen CC, Zborowski P, Ritchie SA, Van Den Hurk AF. Efcacy of bird- tera: Culicidae) bloodmeal sources in Western Australia: implications baited traps placed at diferent heights for collecting ornithophilic mos- for arbovirus transmission. J Med Entomol. 2009;46:1167–75. quitoes in eastern Queensland, Australia. Aust J Entomol. 2009;48:53–9. 25. Neuwirth E. Package ‘RColorBrewer’. 2015. http://www.colorbrewe 48. Johnston E, Weinstein P, Slaney D, Flies AS, Fricker S, Williams C. Mosquito r2.org. communities with trap height and urban-rural gradient in Adelaide, Stephenson et al. Parasites Vectors (2019) 12:156 Page 11 of 11
South Australia: implications for disease vector surveillance. J Vect Ecol. 59. Scott TW, Amerasinghe PH, Morrison AC, Lorenz LH, Clark GG, Strick- 2014;39:48–55. man D, et al. Longitudinal studies of Aedes aegypti (Diptera: Culicidae) 49. Jansen CC, Webb CE, Northill JA, Ritchie SA, Russell RC, van den Hurk AF. in Thailand and Puerto Rico: blood feeding frequency. J Med Entomol. Vector competence of Australian mosquito species for a North American 2000;37:89–101. strain of West Nile virus. Vector Borne Zoonotic Dis. 2008;8:805–11. 60. Scott TW, Chow E, Strickman D, Kittayapong P, Wirtz RA, Lorenz LH, 50. Kay BH, Hearnden MN, Oliveira NM, Sellner IN, Hall RA. Alphavirus Edman JD. Blood-feeding patterns of Aedes aegypti (Diptera: Culicidae) infection in mosquitoes at the Ross River reservoir, north Queensland, collected in a rural Thai village. J Med Entomol. 1993;30:922–7. 1990–1993. J Am Mosq Cont Assoc. 1996;12:421–8. 61. Calvez E, Guillaumot L, Girault D, Richard V, O’Connor O, Paoaafaite T, 51. Williams CR, Kokkinn MJ, Smith BP. Intraspecifc variation in odor-medi- et al. Dengue-1 virus and vector competence of Aedes aegypti (Diptera: ated host preference of the mosquito Culex annulirostris. J Chem Ecol. Culicidae) populations from New Caledonia. Parasit Vectors. 2017;10:381. 2003;29:1889–903. 62. Brady OJ, Golding N, Pigott DM, Kraemer MU, Messina JP, Reiner RC Jr, 52. Standfast H, Barrow G. Studies of the epidemiology of arthropod-borne et al. Global temperature constraints on Aedes aegypti and Ae. albopictus virus infections at Mitchell River Mission, Cape York Peninsula, North persistence and competence for dengue virus transmission. Parasit Vec- Queensland. Trans R Soc Trop Med Hyg. 1968;62:418–29. tors. 2014;7:338. 53. Kilpatrick AM, Kramer LD, Campbell SR, Alleyne EO, Dobson AP, Daszak P. 63. Jansen CC, Williams CR, van den Hurk AF. The usual suspects: compari- West Nile virus risk assessment and the bridge vector paradigm. Emerg son of the relative roles of potential urban chikungunya virus vectors in Infect Dis. 2005;11:425. Australia. PLoS One. 2015;10:e0134975. 54. Hongoh V, Berrang-Ford L, Ogden NH, Lindsay R, Scott ME, Artsob H. A 64. González-Salazar C, Stephens CR, Sánchez-Cordero V. Predicting the review of environmental determinants and risk factors for avian-associ- potential role of non-human hosts in Zika virus maintenance. EcoHealth. ated mosquito arboviruses in Canada. Biodiversity. 2009;10:83–91. 2017;14:171–7. 55. Andreadis TG. The contribution of Culex pipiens complex mosquitoes to 65. Chahad-Ehlers S, Fushita AT, Lacorte GA, de Assis PC, Del Lama SN. transmission and persistence of West Nile virus in North America. J Am Efects of habitat suitability for vectors, environmental factors and host Mosq Cont Assoc. 2012;28:137–51. characteristics on the spatial distribution of the diversity and prevalence 56. Jacups SP, Carter J, Kurucz N, McDonnell J, Whelan PI. Determining of haemosporidians in waterbirds from three Brazilian wetlands. Parasit meteorological drivers of salt marsh mosquito peaks in tropical northern Vectors. 2018;11:276. Australia. J Vect Ecol. 2015;40:277–81. 66. Gard G, Marshall ID, Woodroof GM. Annually recurrent epidemic polyar- 57. Hu W, Mengersen K, Dale P, Tong S. Diference in mosquito species thritis and Ross River virus activity in a coastal area of New South Wales. II. (Diptera: Culicidae) and the transmission of Ross River Virus between Mosquitoes, viruses and wildlife. Am J Trop Med Hyg. 1973;22:551–60. coastline and Inland Areas in Brisbane, Australia. Environ Entomol. 2010;39:88–97. 58. Stephenson EB, Peel AJ, Reid SA, Jansen CC, McCallum H. The non-human reservoirs of Ross River virus: a systematic review of the evidence. Parasit Vectors. 2018;11:188.
Ready to submit your research ? Choose BMC and benefit from:
• fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year
At BMC, research is always in progress.
Learn more biomedcentral.com/submissions
Appendix B: RRV field survey protocol
This protocol was developed during this PhD as part of planning for the Brisbane field studies of Objective 4. It outlines in detail the rationale and methodology employed for this part of the research. This document was written by Amanda Murphy with input from Eloise Skinner (formerly Stephenson) of Griffith University.
Chapter 9: Appendices 260
Field study protocol: Investigating Ross River virus vectors and hosts in Brisbane
Contents
1. Investigators and institutions 262 2. Background 263 3. Study aims and objectives 264 4. Research plan and methodology 265 4.1. Study sites 4.2. Survey and sampling strategy 4.2.1. Flora surveys 4.2.2. Fauna surveys 4.2.3. Mosquito surveys 4.2.4. Climatic conditions 4.3. Laboratory analysis methods 4.3.1. Identification of mosquito species 4.3.2. Assessment of disease prevalence in mosquitoes 4.4. Data management and analysis 5. Resources and funding required 275 6. Coordination and communication 275 7. Outcomes and significance 276 8. Study timeline 278 9. Reference list 278 10. Appendices 279
Chapter 9: Appendices 261
1. Investigators and institutions This field study is a collaboration between QIMR Berghofer Medical Research Institute (QIMRB), Griffith University, the Queensland University of Technology (QUT), and the Queensland Department of Health (QH). The study will be overseen by a multidisciplinary team of experts, spanning the fields of ecology, wildlife biology, entomology and disease epidemiology, and will be coordinated by two PhD Candidates. The study will also be supported by the efforts of a number of volunteer field data collectors. The study Investigators and their roles are listed below:
Fieldwork Coordinators
Ms Eloise Stephenson, PhD Candidate, Disease Ecology Group, Griffith University
Ms Amanda Murphy, PhD Candidate, Mosquito Control Laboratory, QIMRB & QUT
Research Advisors
Prof. Hamish McCallum, Disease Ecology Group, Griffith University
Dr. Alison Peel, Disease Ecology Group, Griffith University
Prof. Simon Reid, School of Population Health, University of Queensland
Dr. Cassie Jansen, Queensland Department of Health
A/Prof. Gregor Devine, Mosquito Control Laboratory, QIMRB
Dr. Francesca Frentiu, School of Biomedical Sciences, QUT
A/Prof. Wenbiao Hu, School of Public Health & Social Work, QUT
Prof. Louise Hafner, School of Biomedical Sciences, QUT
Chapter 9: Appendices 262
2. Background Ross River virus (RRV) is responsible for the most widespread and frequently reported mosquito‐borne disease in Australia [1, 2]. The disease symptoms range from rash, fever and fatigue to severely debilitating effects such as polyarthritis. The virus is maintained in complex reproductive cycles between mosquitoes, humans and animals. The disease has significant public health and economic impact in Australia, particularly in Queensland where the disease rates are high [3].
RRV is maintained in the environment through enzootic cycles between the vectors (mosquitoes) and wildlife reservoirs, with subsequent transmission (‘spill over’) into human populations. More than 30 species of mosquitoes are capable of transmitting RRV, and antibodies to the virus have been found in several vertebrate species including kangaroos, possums, humans, dogs, cats and horses, suggesting a complex reservoir‐vector interface [4]. During the past 20 years, increasing numbers of cases have been observed within metropolitan centres. This may have been influenced by urban expansion in proximity to wildlife reservoir and mosquito vector habitats, though the specific risks are unclear [1, 5].
At present, there is no treatment or vaccine available for RRV. Management of the disease is limited to controlling vector populations. Whilst this has been found to be effective at minimising the number of human cases, it is not enough to prevent outbreaks of RRV [6]. This was particularly evident in 2015, which saw the largest outbreak of RRV across Australia in over 30 years. More than 9,500 human notifications were reported, almost double the annual average [3]. Sixty‐five percent of cases (6,192) were notified from Queensland, with around half from metropolitan Brisbane – a five‐fold increase compared with the previous 4 years. This outbreak highlighted the ongoing public health and economic importance of this disease, as well as the need to understand the specific determinants of RRV transmission.
A lack of a comprehensive understanding of the transmission dynamics of RRV limits the ability to predict and manage future outbreaks. This study seeks to investigate the underlying determinants of contemporary RRV epidemiology in urban
Chapter 9: Appendices 263
and peri‐urban locations of Brisbane, and to apply this understanding to improved management strategies for RRV. Its focus will be on quantifying the relationships between environmental and ecological factors and the burden of RRV disease in humans.
3. Study aims and objectives This broad aim of this study is to quantify ecological factors influencing RRV transmission within different potential (vector and vertebrate host) habitat types in Brisbane. Specific objectives are to: a) Measure seasonal changes in flora within a variety of localised environments in Brisbane, b) Measure seasonal changes in fauna diversity and abundance within the same environments, c) Measure seasonal change in mosquito species abundance and diversity within the same environments, d) Investigate relationships between vectors and hosts in these environments to incriminate specific vectors & hosts in RRV transmission. This will involve two approaches: i. Quantify host species availability and abundance relative to mosquito species presence and abundance, within given habitats; and ii. Investigate relationships between vector‐host interactions and RRV transmission through the assessment of mosquito feeding behaviours, relative to available host species in the same habitats. e) Assess whether environmental, vertebrate host and entomological factors within different (urban/peri‐urban/rural and coastal/inland) habitat types are related to patterns of reported human RRV disease.
Chapter 9: Appendices 264
4. Research plan and methodology In South East Queensland, where there is a high incidence of RRV disease, there is potential to investigate some of the key factors impacting the expansion of RRV disease. There is a range of ecological environments where the interactions of humans, mosquitoes and reservoir host species can be explored. The proposed research will seek to understand the epidemiology of RRV disease within localised urban and peri‐urban environments in the Brisbane region, shown in Figure 1.
Figure 1. Local government geographical boundaries of Brisbane City.
Source: http://results.ecq.qld.gov.au/interactive_maps/local‐government‐ map.html?council=Brisbane%20City%20Council
Chapter 9: Appendices 265
4.1 Study Sites
Nine separate collection sites within urban and peri‐urban Brisbane environments were selected, which represent habitats where key vector and mammalian host species might intersect. The study sites are shown in Figure 2. Sites were selected based on possessing a variety of key characteristics relevant to RRV transmission, summarised in Table 1.
Six of the sites are existing locations for mosquito surveillance activities by the Brisbane City Council, and where historic mosquito distribution data exists. Mosquito trapping activities in these sites will be conducted in partnership with Brisbane City Council staff. Three additional sites were chosen with urban and peri‐urban characteristics not already covered by the other six sites. Briefly, key criteria to define potential sites were:
Recent high or low rates of human RRV notifications (previous 3 years); High versus low human population density; Varied geo‐ecological environments (different habitats for suspect hosts and vectors); e.g. coastal/inland, urban/suburban/rural, north/south of the river; and Presence of priority vector mosquitoes and priority reservoir host species.
Chapter 9: Appendices 266
a)
b)
Figure 2. Google Earth maps of the study sites. a) Map of the nine selected study sites, six existing surveillance sites monitored by Brisbane City Council (yellow pins), plus three additional sites (red pins). b) Example of 4‐plot layout per site (Pullenvale site).
Chapter 9: Appendices 267
Table 1. The nine selected study sites, and characteristics of each.
Human Human pop. RRV Recent RRV Mosquito SLA name Site Geography Ecosystem(s) Animal diversity pop. size density cases rates* diversity Private property & 19.92 north metropolitan, 16,741 82/100,000 birds, kangaroos, Bracken Ridge Bracken Ridge people/ha 41 mainly saltwater coastal wetland (high) (med) horses Reserve (med) 16.85 Herbert St Park & south metropolitan, 3,401 59/100,000 birds, bats, mainly saltwater, Lota people/ha 6 Whites Rd Park coastal wetland (low) (med) possums some freshwater (med) metropolitan, 17.3 5,883 85/100,000 birds, bats, mainly saltwater, Chermside Raven St Reserve north riparian bushland, people/ha 15 (med) (high) wallabies some freshwater West dry eucalypt forest (med) 23.44 mainly Wishart metropolitan, 11,238 24/100,000 birds, bats, south people/ha 8 freshwater, Wishart Community Park parkland (high) (low) wallabies (med) unique species 17.04 mainly metropolitan, 12,800 34/100,000 Indooroopilly Robertson Park south‐west people/ha 13 birds, bats freshwater, some parkland (high) (low) (med) saltwater 17.63 mainly metropolitan, 5,084 105/100,000 Cliveden Ave south‐west people/ha 16 birds, horses freshwater, Corinda parkland (med) (high) (med) unique species 36.48 inner urban, 6,194 81/100,000 birds, bats, mainly Red Hill Woolcock Park people/ha 15 north parkland (med) (high) possums freshwater (high) 12.73 inner urban, 2,131 31/100,000 birds, bats, mix of fresh and Herston Gould Park people/ha 2 north parkland (low) (low) possums salt (low) rural, 0.94 birds, wallabies, Pullenvale Forest 2,804 166/100,000 mainly Pullenvale west riparian bushland, people/ha 14 possums, horses, Park (low) (high) freshwater dry eucalypt forest (low) deer * RRV infection rate calculated based on number of RRV notifications reported during the previous 3‐year period (2014‐2016) relative to the human population density of the suburb (SLA).
Chapter 9: Appendices 268
4.2. Survey and sampling strategy The surveys and sampling will be led by the two fieldwork coordinators, who will oversee and guide a team of student volunteers who will assist with the data collection (a copy of the volunteer sign up sheet can be found in Appendix 10.1). The methodologies for this fieldwork are adapted from similar research undertaken by Kilpatrick et al. (2006) for West Nile virus. The methodology comprises surveys and sampling of flora, fauna and mosquitoes in diverse habitats across Brisbane (described above in the Study sites section) over an eight‐month period.
For each of the nine study sites, up to 4 plot areas will be measured/marked, each measuring 100m x 100m with at least 50m between them. Within those plots, the following survey and sampling methods will be conducted for each ecological component:
4.2.1 Flora surveys
Flora surveys will be conducted in each site, once per month, to observe any seasonal changes that may impact movement/behaviour of fauna species. These will include:
• An initial survey of the flora species and density in each site (September 2017). • Density will be measured using the T‐square method (as described within the Queensland Department of Environment and Heritage Protection Flora Survey Guidelines). • For each fauna survey, notes will be made on the flowering state of any plants.
4.2.2 Fauna surveys
Three fauna surveys will be conducted in each of the 100 x 100m plots; bird surveys, mammal surveys and scat surveys. These surveys will be undertaken twice a month, ideally spaced at least 7 days apart. Surveys will comprise:
Chapter 9: Appendices 269
Bird surveys
• Bird surveys will commence 30 mins before dawn, and will comprise 30 person‐minutes per 100 x 100m plot. • Each survey involves the observer walking slowly and quietly through the site for five minutes, looking and listening for birds, taking a different path on each occasion. • Birds seen or heard within the site are counted and recorded on a data sheet, avoiding counts of the same individuals more than once in each survey. Any animals observed will also be noted. • To vary the survey times at each site, surveys in each plot should, where practical, be conducted in a different order on each survey visit.
Mammal surveys
• Mammal surveys will commence 30 minutes after sunset, with the aim to observe nocturnal activity of mammals and marsupials in the area. • Each survey involves an observer/s walking slowly and systematically through the each of four 100 x 100m plots within the survey site, and searching trees, ground and fence lines using head torches and hand‐held spotlight torches. • Within each 100 x 100m plot, 2 x 30 person‐minute searches will be undertaken. More than one observer can search the site at the one time (e.g. two observers search the site for 15 minutes). • If the site is lacking woody vegetation (e.g. a grassland, a cleared paddock or an arid regional ecosystem), then it is acceptable to reduce the effort spent spotlighting the site. In these situations, a 10‐person minute search is sufficient. It is important to note this on the datasheet. • Animals seen or heard can be recorded as on‐site (within the 100 x 100 m area), near site (within 50 m of the boundary of the site) or offsite (> 50 m from the boundary of the site). Any birds observed will also be noted.
Chapter 9: Appendices 270
Scat surveys
• Scat surveys will be conducted during the hour after sunrise and/or the hour before sunset. Scats will be noted within these time period for each 100 x 100m plot.
4.2.3 Mosquito surveys
Mosquito surveys will take place at least twice/month, across at least two mornings and evenings, spaced at least 7 days apart. Surveys will be performed between at least one hour before sunset and one hour after sunrise, and will include a variety of trapping methods to maximise the chance of capturing diverse species, to ensure accurate characterisation of species abundance, as well as the highest possible yield of blood fed mosquitoes:
• Mosquito traps will be set at least one hour before sunset, and collected at least one hour after sunrise. For each site, traps will be spaced at least 100m apart, with different combinations of trap types used in different plot areas. Traps will include: o 2 x CO2/octenol‐baited light traps at 1.5m height (possibly with FTA cards) o 4 x Biogents’ Gravid Aedes Traps (BG‐GAT) + leucerne mix, placed at ground level • Vegetation and resting surfaces (such as bridges or building structures and/or resting boxes) in at least 2 100m x 100m plots/site will be aspirated to capture resting mosquitoes, for 10 person‐minutes/plot. Aspiration will take place 1 hour after sunrise, and/or 1 hour before sunset. • Where possible, larval surveys of mosquito breeding sites will opportunistically be conducted for each site. Larvae will be laboratory reared and identified to species.
Chapter 9: Appendices 271
4.2.4 Climatic conditions Temperature, relative humidity, wind speed/direction, recent tidal levels and rainfall will be recorded for each survey, using both: • Broad measurements obtained from the Australian Bureau of Meteorology, and/or • Specific local measurements available from Brisbane City Council weather gauges, and/or • Specific local measurements obtained using a Kestrel handheld weather meter.
4.3 Laboratory analysis methods
4.3.1 Identification of mosquito species All captured mosquitoes will be sorted using a chilled table, and identified to species level using a dissection microscope, and counted. Blood fed mosquitoes identified will be separated, stored at ‐80◦C, and processed according to Section 4.3.2. Non‐blood fed mosquitoes will be stored at ‐20◦C until the time of viral analysis.
4.3.2 Assessment of disease prevalence in mosquitoes Sugar‐baited viral detection cards and/or PCR methods Sugar‐baited viral detection cards (e.g. FTA cards) will be placed in light traps used in the field collections (Hall‐Mendelin et al. 2010). For traps with cards positive for virus, mosquitoes from that trap will be further assessed to identify the mosquito species host of the virus. Pools of mosquito species (between 10 and 50 individuals) from each survey will be tested for RRV RNA by RT‐PCR, as described by Ritchie et al. 1997.
Laboratory analyses of mosquito blood meals All captured mosquitoes will be identified to species level, and counted. Blood fed mosquitoes identified will be separated and classified according to their
Chapter 9: Appendices 272
engorgement status. Captured blood‐fed mosquitoes will be separated by location and species, and stored at ‐80◦C until the time of analysis. Their blood meals will be analysed to assess: a) the host species identity i.e. the source of the blood meal; and b) the presence of RRV antibodies within the blood meal.
Briefly, mosquito abdomens are removed, homogenized in phosphate‐ buffered saline (PBS), centrifuged and the supernatant removed. The host species identity of the blood meal is determined through DNA extraction from the supernatant and PCR amplification of the cytochrome b segment of mitochondrial DNA (Ngo & Kramer 2003), using broad species‐specific primers (e.g. mammalian, avian, marsupial). PCR products are then electrophoresed, purified and sequenced to species level [7]. Nucleotide sequencing is then performed on the amplified product. Feeding preference for each mosquito species (Pi) will be calculated using:
Pi = fraction of total blood meals from host i (density of species i / total density of available host species)
If mosquitoes feed host species proportionally to their abundance, then Pi will equal 1.
For identification of RRV antibodies within the mosquito blood meal (i.e. in the blood of the host species), a xenodiagnostic technique will be used. This method is currently being developed at QIMRB, and detects RRV antisera within the mosquito blood meal by microPRNT method (manuscript in preparation).
4.4 Data management and analysis
Data management and record keeping Environmental, flora, fauna and bird observation data will be collected using forms adapted from the Terrestrial Vertebrate Fauna Survey Guidelines for Queensland. Data fields will be populated by fieldwork coordinators and volunteers and ultimately fed into MS Excel spreadsheets. An example of the data collection
Chapter 9: Appendices 273
(Excel) sheet is attached in Appendix 10.2. Briefly, minimum data recorded will comprise: • Observation date/time and name of surveyors • Location description and coordinates • Number and taxon name of vertebrate species observed • Description of flora/vegetation type and state at the time of observation (e.g. flowering plants) • Environmental/climate conditions at, and preceding, the time of observation (e.g. rainfall, tides, wind speed/direction) • Time/effort of surveys conducted per site
Data generated from mosquito trapping will be entered into a separate data collection (Excel) sheet following laboratory sorting and identification of numbers of each species caught via each trapping method. This data collection sheet can be found in Appendix 10.3. All electronic data will be password‐protected and only investigators directly involved in the study will have access to this data. As per institutional policies for research, data will be retained for a minimum of 5 years.
Statistical Analyses Data will be analysed using SPSS and/or R statistical software. Categorical variables will be compared using the Chi‐squared test. Proportions will be compared using the Wilcoxon or Mann Whitney test. Spearman’s correlation coefficient will be used to analyse associations between variables. Multivariate analyses of spatiotemporal and seasonal relationships between reservoir host and vector species, climate, environmental factors, human population density and RRV infection rates will be conducted using Poisson autoregressive models developed at QUT, as well as adapted methods for calculation of reproductive ratio used by Kilpatrick et al. 2006 [8, 9].
Chapter 9: Appendices 274
5. Resources and funding required Field supplies and other resources required for this project will be provided by Griffith University, QUT, QIMRB and Queensland Department of Health. Resources contributed by each organisation are summarised in Appendix 10.4. The two field work coordinators are supported by Australian Postgraduate Awards from the Australian Government, and student research allocation funding from their respective Universities. Their insurance cover is provided by their Universities. Volunteers will be recruited through Griffith University and covered by Griffith’s insurance. Analysis of the data will require statistical analysis and GIS mapping software (SPSS, R, ArcGIS), which are available through Griffith and through the School of Public Health and Social Work at QUT. Interpretation of data collected will be aided by consultation with public health and entomological expertise at Queensland Health and Brisbane City Council; however, no direct costs are involved.
6. Coordination and communication The two fieldwork coordinators will coordinate communications with volunteers, study collaborators and research advisors, and members of the public. Although both coordinators will assist in communication with all parties as/when required, Ms Stephenson will be the primary contact for volunteer communications, while Ms Murphy will primarily lead communication with public organisations. Their contact details are as follows:
Ms Eloise Stephenson (Volunteer communications) E‐mail: [email protected]
Ms Amanda Murphy (Public communications) E‐mail: [email protected]
Chapter 9: Appendices 275
Because some of the study sites are located in densely population residential areas, there are risks that residents will observe, interfere with, and/or be concerned about study activities. To mitigate these risks, the following communication strategies will be employed:
1. Distribution of general information sheets to inform residents of our activities: In the two study sites with a high density human population (Highgate Hill and Red Hill), residents in the surrounding areas will be informed of the study activities taking place, in particular, the presence of research staff with torches and binoculars during the early morning and evening hours. The information notices to be distributed in these areas can be found in Appendix 10.5. 2. Signs on mosquito traps: In all sites, placement of overnight mosquito traps will have signs attached advising of the purpose of the traps and contact information for any public concerns. 3. Informing Brisbane City Council staff: Local council offices responsible for overseeing the public spaces where study activities will be conducted will be notified by phone and in writing in advance of the study activities taking place. 4. Documenting permission to enter private land: In the case that mosquito trapping activities require entrance to privately owned premises or land, an informed consent process will be followed, developed in liaison with the QIMRB Human Research Ethics Committee (HREC). The private landowner will be provided with a study information sheet and will be asked to sign a consent form (see Appendix 10.6).
7. Outcomes and significance RRV disease contributes a significant burden of morbidity and mortality in Australia, particularly in Queensland. RRV also has significant economic costs to the Australian health system and economy, and incidence is rising annually. However,
Chapter 9: Appendices 276
few specific prevention or control strategies exist. This study will provide information on the underlying causes of RRV disease outbreaks. This information can inform preventative efforts of public health departments in SEQ, and possibly in other areas of Australia and internationally. The outcomes of the study may also inform the direction of future research for this disease. Key points about the study outcomes: • This study aims to address central unresolved questions relating to transmission of Australia’s most prevalent mosquito‐borne disease – specifically, the most important mosquitoes, host species and environments mediating ongoing RRV transmission. • This research may inform wildlife conservation and mosquito borne disease control efforts in Australia and internationally, particularly in Pacific Island Countries where RRV is also endemic. • It will be the first field study to conduct a broad, multidisciplinary assessment of risk factors contributing to the burden of Ross River virus disease.
Specific research outcomes of this study may include peer‐reviewed publication of: • Seasonal abundance and diversity of flora and fauna within different environments of Brisbane • Seasonal abundance and diversity of mosquito species within different environments of Brisbane • Association between proximity to flora/fauna and RRV disease risk in Brisbane • Association between mosquito species abundance/diversity and RRV disease risk in Brisbane • Predictive model of potential ecological and environmental drivers of RRV disease risk in Brisbane (mosquitoes, hosts, people, environment/ climate).
Chapter 9: Appendices 277
8. Study timeline This study will be conducted over a 9‐month period, between October 2017 and August 2018, as indicated in the table below. Preliminary assessment of site suitability will take place during September 2017. The field data collection period will encompass the Spring, Summer and Autumn seasons in Brisbane, capturing the period preceding and during the typical RRV peak notification period: on average between February and May each year. Conducting the field study over this time period aims to capture the seasonal ecological change that may influence RRV outbreaks. Following the field data collection work, laboratory and statistical analyses will be conducted to assess the trends and associations found, and the results reported and disseminated.
Time period (years/months)
2017 2018 Study Phase Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug 1 Site selection and training 2 Field data collection Lab and computer data 3 analyses 4 Reporting study results
9. Reference List 1. Russell, R.C., Ross River virus: ecology and distribution. Annu Rev Entomol, 2002. 47: p. 1‐31. 2. Harley, D., A. Sleigh, and S. Ritchie, Ross River virus transmission, infection, and disease: a cross‐disciplinary review. Clin Microbiol Rev, 2001. 14(4): p. 909‐32, table of contents. 3. Government, A., National Notifiable Diseases Surveillance System. 2016, Commonwealth of Australia.
Chapter 9: Appendices 278
4. Claflin, S.B. and C.E. Webb, Ross River Virus: Many Vectors and Unusual Hosts Make for an Unpredictable Pathogen. PLoS Pathog, 2015. 11(9): p. e1005070. 5. Jardine, A., P.J. Neville, and M.D. Lindsay, Proximity to mosquito breeding habitat and Ross River virus risk in the Peel region of Western Australia. Vector Borne Zoonotic Dis, 2015. 15(2): p. 141‐6. 6. Tomerini, D.M., P.E. Dale, and N. Sipe, Does mosquito control have an effect on mosquito‐borne disease? The case of Ross River virus disease and mosquito management in Queensland, Australia. J Am Mosq Control Assoc, 2011. 27(1): p. 39‐44. 7. Jansen, C.C., et al., Blood sources of mosquitoes collected from urban and peri‐ urban environments in eastern Australia with species‐specific molecular analysis of avian blood meals. Am J Trop Med Hyg, 2009. 81(5): p. 849‐57. 8. Hu, W., et al., Exploratory spatial analysis of social and environmental factors associated with the incidence of Ross River virus in Brisbane, Australia. Am J Trop Med Hyg, 2007. 76(5): p. 814‐9. 9. Hu, W., et al., Difference in mosquito species (Diptera: Culicidae) and the transmission of Ross River virus between coastline and inland areas in Brisbane, Australia. Environ Entomol, 2010. 39(1): p. 88‐97.
10. Appendices 10.1 Volunteer sign up sheet 10.2 Data collection sheets: flora, fauna and environmental variables 10.3 Data collection sheet: record of mosquito species collected per trapping method 10.4 List of study resources provided by each institution 10.5 Resident information sheet 10.6 Consent form for activities on private property
Chapter 9: Appendices 279
Appendix 10.1 (pdf) Appendix 10.2 (3 x pdfs)
Chapter 9: Appendices 280
Appendix 10.3: Mosquito trapping record General information Date: BRKN LOTA PLNV GLDP CHRM WISH REDH Study site (code): INDR CRND OTHER:______Surveyor name:
Survey start time:
Survey end time:
Climate/environment characteristics Cloud cover level
(none, low, med, high): Moon phase:
(zero, quarter, half, full): Temperature (deg): Wind speed/direction
(calm, light, moderate, strong): Relative humidity:
Precipitation:
Other:
Flora characteristics
Vegetation type(s) present:
Vegetation status
(flowering, abscission etc.): Flower abundance
(none, low, med, high):
Other comments:
Other notes: ______
______
______
Chapter 9: Appendices 281
Mosquito trapping record Site:______Date (when trap set): ______
Plot Trap Trapping Trap location (tree, creek, Trap Trap start Trap finish Adult Other comments: Mozzies no. no. method (LT, under bridge) height time date/ time mosquitoes (tree coverage or open area, grass caught GAT, ASP, BG) (GR, observed? or leaf GR cover, creek status, (Y/N)? 1.5m) (Y/N) climate status)
Trapping method = LT (CO2/octenol baited light trap), GAT (Gravid Aedes Trap), ASP (aspiration), BG (Biogents’ Sentinel Trap) Trapping location = NW (near water/creek), S (in shrubs), UB (under bridge), NL (near log), GR (on ground), DR (in drain) O/S (off site: outside of 100m x 100m plot)
Notes: ______
______
Chapter 9: Appendices 282
Larval sampling record Site:______Date (when sampled): ______
Plot Close Breeding habitat GPS Size of pool Water type Vegetation Number Number of Other comments: no. to type(tree, creek, coordinate (fresh/brackish/ present of dips larvae/ (tree coverage or open area, trap drain, under bridge) s polluted, (floating, pupae per acquatic fauna, creek status) no? clear/cloudy/ algae, dip stagnant, sunlit/ emergent)? shaded)
Notes: ______
______
Chapter 9: Appendices 283
Appendix 10.4 List of RRV Fieldwork resources provided by the respective institutions. Institution Resources Queensland provided Griffith QIMRB QUT Student’s own cost Health General ‐ Field vehicles ‐ 2 x Hand held GPS ‐ Flagging tape ‐ 4 x 100m tape measures ‐ camera ‐ extra torches
‐ pens/sharpies ‐ sunscreen, mosquito repellent ‐ data collection forms (flora, fauna, bird, environment, mosquitoes – adult + larvae) ‐ site maps ‐ study protocol Flora surveys ‐ 2 x Diameter at breast height tape measures
‐ tree and plant ID keys ‐ Ziplock bags Fauna surveys Spotlighting ‐ 4 x binoculars ‐ 6 x head torches ‐ 4 x spotlights ‐ mammal ID book
Chapter 9: Appendices 284
Institution Resources Queensland provided Griffith QIMRB QUT Student’s own cost Health Scat ID ‐ plastic gloves ‐ ziplock bags ‐ sharpie ‐ scat ID book Bird surveys ‐ 4 x binoculars ‐ Voice recorders ‐ 2 x Bird ID book ‐ ipad or smart phone with
‐ countdown timer the Australian birds app (the app is $30) Mosquito ‐ 3 Light traps – plus ‐ 2 BG traps, ‐ surveys dry ice, 3 batteries collection bags, 2 ‐ 2 aspirators, batteries collection cups, 2 ‐ 6 gravid traps, plus batteries water/attractant ‐ esky/ice to store for them, 6 aspirated batteries? mosquitoes ‐ larval dippers/collection tubes (periodic use only) ‐ data loggers for measuring breeding
Chapter 9: Appendices 285
Institution Resources Queensland provided Griffith QIMRB QUT Student’s own cost Health site parameters (periodic use only) Laboratory ‐ PCR reagents analyses ‐ microPRNT reagents Statistical ‐ Statistical & spatial analysis software & expertise (SPSS, R, ArcMap) analyses
Chapter 9: Appendices 286
Appendix 10.5 Resident information sheet
18th September 2017
Dear Resident,
We are writing to advise you of planned research activities to be undertaken in your neighbourhood between October 2017 and May 2018.
This will involve bi‐monthly surveys of birds, nocturnal mammals and mosquitoes present in this area (2 mornings and 2 evenings each month). The purpose of this research is to collect accurate information about the diversity and abundance of wildlife and mosquito species around Brisbane. This information will be used to inform conservation as well as mosquito‐ borne disease prevention efforts.
The best time for us to observe activity of local species is in the early hours of the morning (from 5‐8am) and in the early evening hours (between 5‐9pm). During these hours, you may notice our team members setting or retrieving mosquito traps in the area, and also searching the trees/fence lines with torches. As these surveys require us to work quietly, and with minimal impact on the species present, we hope there will be little to no disturbance for you.
If you have any questions about the research, or would like to know the specific dates we will be in the area, please don’t hesitate to contact us. Please also feel free to come and say hello if you see us in your neighbourhood.
Sincerely,
Eloise Stephenson Amanda Murphy Principal Investigator Principal Investigator Griffith University Queensland University of Technology & QIMRB Mobile: XXXX XXX XXX Mobile: XXXX XXX XXX e‐mail: [email protected] [email protected]
Chapter 9: Appendices 287
Appendix 10.6: Consent form for activities on private property
Participant consent form
THIS IS YOUR CONSENT TO PARTICIPATE IN A RESEARCH STUDY
Research Study P2238 ‐ Investigating Ross River virus vectors and hosts in Brisbane.
Declaration by Participant
By signing below, I confirm the following (tick appropriate):
I have been given oral and written information for the above study and have read and understood the information given, or someone has read it to me in a language that I understand. I have had sufficient time to consider participation in the study and have had the opportunity to ask questions, and I am satisfied with the answers I have received. I understand that my participation is voluntary and I can at any time freely withdraw from the study without giving a reason and there will be no action taken against me. I understand the purposes, procedures and risks of the research described in the project, and I freely give my informed consent to participate under the conditions stated. I understand that I can discuss any concerns related to my participation in the study with the principal investigators. I understand that any/all of my personal details shared with researchers will be treated as STRICTLY CONFIDENTIAL and will not be published in any scientific articles or other reports. I understand that I will be given a signed copy of this document to keep. Name of Participant (please print) ______
______Address
______
Signature ______
Date ______
Chapter 9: Appendices 288
Appendix C: Associations between Ross River virus infection in humans and vector‐vertebrate community ecology in Brisbane, Australia
This article reports the results of field surveys of vertebrates and mosquitoes in the RRV‐endemic city of Brisbane, Australia. It focused on assessing the effect of vector and host community structure on human RRV notifications.
See the online article here: https://www.liebertpub.com/doi/10.1089/vbz.2019.2585
Chapter 9: Appendices 289 VECTOR-BORNE AND ZOONOTIC DISEASES Volume XX, Number XX, 2020 ª Mary Ann Liebert, Inc. DOI: 10.1089/vbz.2019.2585
Associations Between Ross River Virus Infection in Humans and Vector-Vertebrate Community Ecology in Brisbane, Australia
Eloise B. Skinner,1 Amanda Murphy,2,3 Cassie C. Jansen,4 Martin A. Shivas,5 Hamish McCallum,1 Michael B. Onn,5 Simon A. Reid,6 and Alison J. Peel1
Abstract
Transmission of vector-borne pathogens can vary in complexity from single-vector, single-host systems through to multivector, multihost vertebrate systems. Understanding the dynamics of transmission is important for disease prevention efforts, but is dependent on disentangling complex interactions within coupled natural systems. Ross River virus (RRV) is a multivector multihost pathogen responsible for the greatest number of notified vector-borne pathogen infections in humans in Australia. Current evidence suggests that nonhuman vertebrates are critical for the maintenance and spillover of RRV into mosquito populations. Yet, there is a limited knowledge of which mosquito vector species and amplifying vertebrate host species are most important for transmission of RRV to humans. We conducted field surveys of nonhuman vertebrates and mosquitoes in the RRV endemic city of Brisbane, Australia, to assess the effect of vector and host community structure on human RRV notifications. Six suburbs were selected across a gradient of human disease notification rates. Differences in vertebrate and mosquito compositions were observed across all suburbs. Suburbs with higher RRV notifi- cation rates contained greater vertebrate biomass (dominated by the presence of horses) and higher mosquito abundances. This study suggests that horse–mosquito interactions should be considered in more detail and that vertebrate biomass and mosquito abundance be incorporated into future RRV modeling studies and considered in public health strategies for RRV management.
Keywords: arbovirus ecology, abundance, competence, diversity, mosquito, vertebrate
Introduction Mosquito-borne pathogens contribute significantly to the global burden of infectious disease in both human and animal btaining a detailed understanding of pathogen-host- populations (Hill et al., 2005). Widely distributed across Downloaded by Queensland University of Technology from www.liebertpub.com at 08/17/20. For personal use only. Ovector dynamics within a given environment is partic- Australia, Ross River virus (RRV) is Australia’s most fre- ularly challenging for multivector multihost pathogens. The quently notified mosquito-borne disease in humans (Austra- relative role of different vectors and/or vertebrate reservoir lian Government, 2018). RRV is a multihost multivector species in supporting pathogen circulation is often largely pathogen with a complex ecology: it persists in all Australian unknown under natural conditions. However, an understand- bioregions and climates and has been isolated from more than ing of transmission is essential for the design of effective 40 species of mosquito under field conditions (Russell, 2002). prevention and control strategies (Weinstein, 1997, Janousek Humans are generally not considered important amplifying et al., 2014, Hoffmann et al., 2015, Gardner et al., 2017) and or maintenance hosts of RRV, but may play a role in out- for predicting net effects of zoonotic mosquito-borne diseases breaks (Aaskov et al., 1981, Harley et al., 2001). However on human populations. specific epidemiological studies quantifying the role of
1Environmental Futures Research Institute, Griffith University, Nathan, Queensland, Australia. 2School of Biomedical Sciences, Queensland University of Technology, Brisbane, Queensland, Australia. 3QIMR Berghofer Medical Research Institute, Mosquito Control Laboratory, Herston, Queensland, Australia. 4Communicable Diseases Branch, Queensland Health, Herston, Queensland, Australia. 5Brisbane City Council, Field Services, Brisbane CBD, Queensland, Australia. 6School of Public Health, The University of Queensland, Herston, Queensland, Australia.
1 2 SKINNER ET AL.
humans in transmitting RRV are limited. Similarly, the rel- identify areas with persistently high or low rates of RRV. At ative roles of various nonhuman vertebrate host species in the time of this investigation, this was the most current da- maintaining transmission of RRV are poorly understood. taset available to estimate long-term trends and select field Strong experimental evidence supports marsupials as sites. Suburbs included had a minimum population number of competent reservoirs (based on viremia profiles which are 1000, and rates were expressed as per 100,000 population, sufficient to infect susceptible vectors), but amplification and relative to the population size of the suburb. Suburbs were transmission competence has also been demonstrated in other represented spatially using the ‘‘State Suburb Codes (SSCs)’’ species groups, including placental mammals and birds (Kay geographical unit. Notifications for each of Brisbane’s SSCs et al., 1986, Stephenson et al., 2018). Due to a myriad of were matched to population data from the 2016 census, factors, experimental infection studies are infrequently un- available from the Australian Bureau of Statistics (1999). dertaken, and further, it is unclear how transmission com- petence of hosts in an experimental setting reflects their Site selection relative role in transmission in natural settings (Claflin and Webb, 2015). As such, many potential reservoir species have Six SSCs were chosen across Brisbane to be included in not yet been investigated. our study (Fig. 1). SSCs were selected to represent: (1) a Although at least 10 mosquito species have been demon- diverse range of human RRV notification rates, (2) varied strated to have some level of vector competence of RRV in geographical and ecological environments (urban, suburban, the laboratory, it is unclear what the relative contributions of coastal, inland), and preferentially (3) locations included in each species are to both enzootic transmission and/or human Brisbane City Council (BCC) mosquito monitoring scheme. disease outbreaks (Russell, 2002, van den Hurk et al., 2010, Study sites were identified within each of these SSCs, as Jansen et al., 2019). It is likely that the vectors associated described in Table 1. with human outbreaks vary across different bioregions of Australia, and this may also be the case with reservoir hosts Vertebrate surveys (Claflin and Webb, 2015). Simply, the key vectors and hosts Within each SSC, three study sites were selected to satisfy associated with the maintenance of RRV, and its spillover two criteria: (1) sites were required to be an accessible public into humans, in different regions of Australia remain unclear. park or nature reserve where it was feasible to establish Of the *5000 human RRV notifications reported in transects of adequate size, and (2) at least one site within each Australia each year, the majority are reported from the state SSC had to include the location of the mosquito trap (selected of Queensland (Australian Government, 2018). Residents of by Brisbane City Council). Each study site comprised a Queensland’s capital city, Brisbane (27.4698 S, 153.0251 E), 100 · 100 m transect (marked using handheld GPS devices; consistently represent the highest proportion of the state’s Garmin-62S) between 50 and 1000 m apart from other sites. cases. In 2014–2015, Queensland experienced the largest Dawn and dusk surveys of nonhuman vertebrates were car- outbreak of RRV ever recorded in Australia, totaling more ried out monthly in each site over a 6-month period that than 6000 human notifications (Jansen et al., 2019, Queens- encompassed the peak RRV transmission season, between land Government, 2019). Brisbane has a subtropical climate, October 2017 and March 2018. which accommodates a diversity of habitats for both vector Methods were adapted from the ‘‘Standard Early Bird and reservoir species. This diversity, in combination with the Search’’ and ‘‘Arboreal Spotlight Search’’ detailed in the persistent high human notification rates of RRV, makes the Queensland Fauna Survey Guidelines (Eyre et al., 2014). In city an ideal location to study RRV disease ecology. brief, dawn surveys commenced 30 min before sunrise and This study aims to assess mosquito and vertebrate host ran for 30 min in each site, sites within a SSC were sampled community ecology across suburbs of Brisbane encompassing sequentially, and the order was mixed at each visit. The locations with a range of RRV notification rates. Specifically, following data were recorded for all nonhuman vertebrate we aim to identify relationships between abundance, diversity, species identified: species, number of individuals, record type and biomass of potential nonhuman vertebrate reservoirs, (i.e., seen, heard, and seen and heard), microhabitat (i.e.,on mosquito vectors, and human notification rates of RRV. the ground, in the canopy, and above the canopy), and whe-
Downloaded by Queensland University of Technology from www.liebertpub.com at 08/17/20. For personal use only. ther the observation was in, near (<50 m), or off (>50 m) the Materials and Methods transect. The same data were also recorded for evening sur- Human notification data veys, which commenced after sunset and were carried out in each transect for 30 person minutes. Equipment used to Long-term human notification data for Brisbane suburbs identify vertebrates included binoculars, handheld torches, were obtained for previous years (2001–2016, inclusive) field guides, and the Birds of Australia mobile device appli- from the Queensland Department of Health (ethics applica- cation (Morcombe, 2003, Knight and Menkhorst, 2010). tion: ref no. P2238). Notifications were based on laboratory evidence of probable or confirmed RRV infection, deter- Mosquito surveys mined through (1) isolation of RRV from patients, (2) de- tection of RRV by nucleic acid testing, or (3) IgG Mosquitoes were collected across the six SSCs in Brisbane seroconversion or a significant increase in IgG antibody level using CDC-stylelight traps (Pacific Biologics, Scarborough, in previously tested patients (Australian Government, 2019), Australia) baited with carbon dioxide (CO2;as*1 kg dry and the suburb of residence of notified cases was recorded. To ice) and 1-octen-3-ol. Trapping was undertaken by BCC as account for high levels of interannual variability in RRV part of its ongoing mosquito surveillance and control pro- notifications (Supplementary Table S1), mean annual notifi- gram. Traps were set once per week before dusk (starting the cation rates from 2001 to 2016 were calculated and used to 5th of September 2017, ending the 26th of March 2018) and VECTOR–HOST COMMUNITY ECOLOGY & HUMAN INFECTION 3 Downloaded by Queensland University of Technology from www.liebertpub.com at 08/17/20. For personal use only.
FIG. 1. (a) Location of Brisbane Local Government Area within Australia, (b) distribution of SSCs in which study sites were located, and from which data were collected. BR, Bracken Ridge; CO, Corinda; CW, Chermside West; IN, In- dooroopilly; LO, Lota; SSC, state suburb code; WI, Wishart. Color images are available online.
collected the following morning. A single trap was located Data analysis within one of the three sites in each SSC used for vertebrate surveys. Traps were set 1.5 m off the ground in proximity to a Variable analysis. Community ecology variables, in- productive larval mosquito habitat. Mosquitoes were identi- cluding biomass (vertebrates only), abundance, and species fied to species level with the aid of morphological keys (Lee diversity, were calculated for vertebrate and mosquito com- et al., 1989, Webb et al., 2016). munities within each site. Species diversity was calculated 4 SKINNER ET AL.
using the Shannon–Weiner index (Boyle et al., 1990), which b accounts for both the abundance and evenness of a given species in a community. Biomass was calculated by multi-
Queensland plying the total number of observed individuals for a given b species by the average adult body mass (kg) of that species and log transformed for analysis. Body mass was derived (2001–2016) from the PANTHERIA dataset for mammals (Jones et al., notification rate
Average annual RRV 2009), and study of Garnett et al. (2015) was used for birds. Generalized linear models (GLM) regression models were used to test whether human notification rate for RRV corre- lated with a given community ecology variable. (2016)
a Community analysis. Vertebrate and mosquito commu- nities within each SSC were analyzed in two ways. First,
Human pop. stacked column charts were constructed to visually represent density community composition. Second, to examine the effect of community compositions on human notification rates, non- metric multidimensional scaling (NMDS) analyses were performed using the vegan R package (Oksanen et al., 2010) (2016)
a and BiodiversityR (Kindt, 2019). These ordinations analyzed 5883 17.3 people/ha 33/100,000 community composition of vertebrates and mosquitoes size Human pop. (separately) for each month at each SSC. The data were auto-transformed and scaled using Bray Curtis dissimilarity indices using the ‘‘metaMDS’’ function. The degree of stress for each NMDS plot was calculated, which indicates the reliability of the outcome, that is, lower stress corresponds with a higher reliability. The ordination of elements was considered arbitrary for stress values of 0.3 or above. The dissimilarity matrices were based on abundances Ecosystem(s) classification for both vertebrates and mosquitoes, and an additional bio-
bushland, dry eucalypt forest mass ordination was generated for vertebrates.
Results Human cases of RRV within each SSC The mean annual notification rate across all Brisbane suburbs was 31 cases/100,000 people between 2001 and 2016. The six SSCs selected for our study exhibited wide variation in the rate of reported RRV (Fig. 2). Of these,
Brisbane City Corinda reported the highest mean annual notification rate,
Geography within followed by Bracken Ridge, Chermside West, Indooroopilly, Australian Bureau of Statistics 2016 census and the RRV notification rate is averaged from 2001 to 2016, from a Wishart, and Lota, respectively. E North, coastal Metropolitan, wetland 16,741 19.92 people/ha 43/100,000 E South, coastal Metropolitan, wetland 3401 16.85 people/ha 17/100,000 E North Metropolitan, riparian E South Metropolitan, parkland 11,238 23.44 people/ha 18/100,000 E South-west, on the river Metropolitan, parkland 12,800 17.04 people/ha 24/100,000 E South-west, on the river Metropolitan, parkland 5084Vertebrate 17.63 people/ha and mosquito 50/100,000 survey summary statistics
Downloaded by Queensland University of Technology from www.liebertpub.com at 08/17/20. For personal use only. Full species lists and counts of individuals observed over the sampling period are summarized in Supplemen- tary Tables S2 and S3. Given the limited number of sites, Ecological, Geographical, and Social Characteristics of Suburbs as Containing Study Sites the power to detect statistically significant results overall S, 153.0340 S, 153.1833 S, 153.0122 S, 153.1000 S, 152.9712 S, 152.9824 1. is limited. However, of the variables tested, logged ver- tebrate biomass had a significant association with hu- man notification rate (GLM: R-squared = 0.11, p = 0.03, Table d.f. = 31). The suburbs reporting the highest human noti- fication rates, Corinda and Bracken Ridge, had the greatest nonhuman vertebrate biomass, which was followed by a decline in biomass in suburbs with lower human notifica- tion rates (Fig. 3a). Total mosquito abundance was also significantly correlated with human notification rate (GLM: R-squared = 0.2271, p = 0.003, d.f. = 31). However, this trend was not linear, with the Bracken Ridge SSC reporting the highest mosquito RRV, Ross River virus; SSC, state suburb code. Human population size and density are derived from the Department of Health data. Bracken Ridge 27.3190 Suburb (SSC) name in which study sites were located Geocoordinates Lota 27.4667 Chermside West 27.3814 Wishart 27.5500 Indooroopilly 27.4984 Corinda 27.5443 abundance and the Wishart SSC reporting the lowest VECTOR–HOST COMMUNITY ECOLOGY & HUMAN INFECTION 5
FIG. 2. Mean annual notification rate (per 100,000 population) for SSCs in Brisbane with a population greater than 1000. SSCs shown in orange are those within which study sites were located, and the dashed black line represents the average notification rate across all SSCs (31/100,000). Color images are available online.
(Fig. 3b). There were statistically significant differences in vertebrate biomass composition (Fig. 5). Specifically, Cor- mosquito diversity, placental mammal abundance, marsupial inda and Bracken Ridge had high total vertebrate biomass, diversity, and marsupial abundance across suburbs, but these almost entirely dominated by Perissodactyla (horses), which significant differences did not correlate with human notifi- was driven by the proximity of horses to a single sampling cation rates (Supplementary Fig. S1c, d, f, g). No significant site. Otherwise, the composition patterns were relatively differences between SSCs nor associations with human no- similar across the remaining four SSCs. Minor differences tification rates were observed for vertebrate abundance and included a larger proportion of marsupials (Diprotodonts) in diversity, bird abundance and diversity, or placental mammal Chermside West and Indooroopilly compared to other SSCs diversity (Supplementary Table S4, Supplementary Fig. S1a, and a greater proportion of avian species (including Psitta- b, e, h, i). ciformes, Passeriformes, and Pelecaniformes) in SSCs with the lowest human notification rates. Overall, the relative Community composition composition of Carnivores was generally consistent across SSCs with mid-range or low human RRV notification rates Across all SSCs mosquito collections comprised mostly (Chermside West, Indooroopilly, Wishart, and Lota) ranging of the following species: Culex annulirostris, Aedes procax, from 33.8% in Indooroopilly to 45.8% in Wishart. Aedes notoscriptus, Aedes vigilax, Anopheles annulipes, Coquillettidia linealis, and Culex sitiens (Fig. 4). There was a 29-fold difference in mosquito abundance between the SSC Ordinations with the highest (Bracken Ridge, n = 13,468) to that of the Dissimilarity matrices resulting from NMDS analyses had
Downloaded by Queensland University of Technology from www.liebertpub.com at 08/17/20. For personal use only. lowest mosquito abundance (Wishart, n = 456) (Fig. 4b, moderately high stress and demonstrated no clear groupings Supplementary Table S2). of suburbs with higher or lower RRV notification rates based SSCs with the highest RRV notification rates were not on vertebrate biomass or mosquito abundance. In general, distinguishable from other SSCs in terms of mosquito species community composition values for each survey were grouped composition. In all SSCs, except Bracken Ridge, both Cx. most closely together by the SSC they were undertaken in annulirostris and Ae. procax contributed >55% of the species (i.e., Bracken Ridge or Indooroopilly), but showed no trend in composition. Compared with other SSCs, Bracken Ridge had grouping between the highest to lowest human notification a unique species composition where Cx. annulirostris and Ae. rate (Fig. 6). procax comprised less than one third, while Ae. vigilax and Cq. linealis dominated (>62% combined). At least 20% of the Discussion species composition in Wishart and Lota (SSCs with the lowest RRV notification rates) were species that have not By integrating data on the abundance, diversity, and bio- previously been considered as candidate vectors of RRV and, mass of potential RRV vertebrate reservoirs and mosquito thus, their vector status is unknown (such as Cx. orbostiensis vectors, we identified associations between high RRV noti- and Ae. vittiger (Harley et al., 2001, Jansen et al., 2019)). fication rates in humans, total mosquito abundance, and total The two SSCs with the highest notification rates were nonhuman vertebrate biomass (particularly due to the pres- distinguishable from all other SSCs based on their nonhuman ence of horses). 6 SKINNER ET AL. Downloaded by Queensland University of Technology from www.liebertpub.com at 08/17/20. For personal use only.
FIG. 3. Boxplots of the two variables with a significant association with human cases (a) nonhuman vertebrate biomass (on a log scale) and (b) mosquito abundance per suburb. Suburbs are ordered from highest human notification rate on the left to lowest on the right. The boxes and whiskers show the extent of variation between months for the relevant variable per suburb.
While variability existed between study sites, the overall greater number of mosquitoes (Lindsay et al., 1996, 1997) or positive relationship observed between total mosquito closer proximity to mosquito habitats (Jardine et al., 2015) abundance and long-term human notification rates of RRV is corresponded with higher human notification rates (Ryan as expected for a mosquito-borne virus. This result is also et al., 1999). Importantly, however, this measure of overall consistent with previous investigations, which found that a mosquito abundance does not necessarily take the vector VECTOR–HOST COMMUNITY ECOLOGY & HUMAN INFECTION 7
FIG. 4. RRV mosquito community composition within each SSC, represented as (a) proportions and (b) totals. SSCs are ordered from highest (left) to lowest (right) RRV notification rate. Total of six trap nights per SSC. Color images are available online.
competence or vectorial capacity of specific species into The dominance of Cq. linealis in Bracken Ridge is also account. Highly competent species at low abundances could interesting as this species is associated with constructed
Downloaded by Queensland University of Technology from www.liebertpub.com at 08/17/20. For personal use only. be important for ongoing transmission. wetlands in urban environments, which has relevance for Although RRV has been isolated from more than 40 authorities managing the increased abundance of this species species of mosquitoes, each differing in general biology, (Russell, 1999, Crocker et al., 2017, Hanford et al., 2019). ecology and, most notably, host feeding patterns (Claflin The dominance of competent vectors at a given site did not and Webb, 2015, Stephenson et al., 2019b), it is likely that always correspond to high notification rates, for example, some species are more important than others as enzootic more than 50% of mosquitoes trapped at Corinda and In- and/or bridge vectors. Further investigations are required to dooroopilly were Cx. annulirostris (a species frequently cited implicate specific mosquitoes as the most important vectors as a vector of RRV), but the two sites have markedly different of RRV in Brisbane; however, we observed notable differ- notification rates. In contrast, Lota, which reported low RRV ences in mosquito communities between suburbs. For ex- notification rates, trapped several mosquito species, which ample, Bracken Ridge had a high dominance of Ae. vigilax have not previously been investigated as vectors of RRV and Cq. linealis whichisofinterestforRRVasthesetwo (no virus isolations) or transmission studies; for example, Cx. species have demonstrated relatively high vector compe- orbostiensis and Ae. vittiger (Harley et al., 2001). While these tence under laboratory investigations (Ryan et al., 2000, species are present in relatively low proportions in these sites, Jefferyetal.,2002)andfrequentlyfeedonbothhumansand the lack of data on their competence and feeding patterns horses (which were abundant in this suburb) (Stephenson makes it difficult to determine their contribution to RRV et al., 2019b). transmission. 8 SKINNER ET AL.
FIG. 5. Nonhuman verte- brate community biomass (kg) categorized by taxo- nomic class within each SSC, represented as (a) propor- tions and (b) totals. SSCs are ordered from highest human notification rate (left)to lowest (right) notification rate. Total of six surveys per SSC. Color images are available online.
The significant positive relationship between vertebrate By comparison, sites with higher notification rates were biomass and RRV notification rate is novel and of interest dominated by a relatively large biomass of horses. Various when mosquito host seeking behavior is considered. Studies studies have suggested that horses are competent amplifiers on mosquito bloodmeals have found that mosquitoes can feed when experimentally infected (developing sufficient viremia preferentially on particular vertebrate species, independent of to infect vectors) (Kay et al., 1987, Stephenson et al., 2018),
Downloaded by Queensland University of Technology from www.liebertpub.com at 08/17/20. For personal use only. their relative abundance (Kilpatrick et al., 2006, Lyimo and that they experience high exposure rates under natural con- Ferguson, 2009, Janousek et al., 2014). When seeking a ditions (Gummow et al., 2018, Stephenson et al., 2019a), and bloodmeal, mosquitoes detect hosts through CO2 and body they have yielded multiple RRV isolates (Azuolas, 1998). heat (Takken and Verhulst, 2013), in addition to numerous The potential role of horses in the maintenance and trans- olfactory factors. Both CO2 and heat emission increase with mission of RRV should be explored further, as their high vertebrate body size (Franz et al., 2010), and thus, larger biomass and interactions with vectors may be important for animals may be more attractive to host-seeking mosquitoes. spillover of RRV to human populations. Biomass accounts for the size of vertebrates, as well as Overall, our findings suggest that vertebrate biomass rather their total abundance. As such, although marsupials are than abundance alone should be assessed in greater detail for considered potential reservoirs of RRV, their biomass con- RRV ecology, but it is most informative when combined with tributed <5% of the community composition in four of the six existing knowledge on the amplification potential (such as SSCs. In SSCs with relatively low RRV notification rates, viremic profile or interactions with vectors) of vertebrate there was a moderate biomass of species that have an un- species. known or limited reservoir potential, including Carnivores The ordinations offered little insight into the disease (namely cats and dogs) and Psittaciformes (namely rainbow ecology of RRV. Specifically, no relevant groupings of sites lorikeets Trichoglossus moluccanus and little corellas Ca- or human notification were ascertained from either the ver- catua sanguinea) (Boyd et al., 2001, Stephenson et al., 2018). tebrate or mosquito communities. One possible explanation VECTOR–HOST COMMUNITY ECOLOGY & HUMAN INFECTION 9
FIG. 6. Results of NMDS analysis for the (a) vertebrate biomass (auto-transformation = square root, stress = 0.1922) and (b) mosquito abundance (auto-transformation = square root, stress = 0.2051). Each point represents a unique monthly survey within each SSC. SSCs are labeled on the right-hand side and colored from highest (dark red)to lowest (light yellow) notifi- cation rate. Color images are available online.
may be that all sites are situated within the same bioregion dilution effect, but if a dilution effect was present, vertebrate (‘‘South Eastern Queensland’’), characterized by similar diversity would have been greatest in areas with low human climatic conditions, with the furthest distance between sites notification rates. We did not observe this. Given the complex *30 km. Although it is assumed that most vertebrates and coupled interactions occurring within this natural system,
Downloaded by Queensland University of Technology from www.liebertpub.com at 08/17/20. For personal use only. mosquitoes did not disperse widely outside of the sites, flying detecting a dilution effect for RRV would require a large- foxes (Tidemann and Nelson, 2004, Roberts et al., 2012), scale study targeting this specific issue. some bird species (Smith and Smith, 2012), and Ae. vigilax Confounding factors within this study include the number (Chapman et al., 1999, Webb and Russell, 2019) readily of replicates, inherent survey method bias, and the choice of travel distances greater than 10 km. In general, while the total sites based on human notifications. First, the limited number community did not vary significantly by SSC within this and purposeful selection of sites and survey months mean that study, it is unlikely that this would be the case if sites with a our findings may not be applicable across Brisbane or to other greater geographic separation were considered, particularly cities. Despite confounding factors within this study, this since RRV has a nationwide distribution, which supports research offers a novel approach to studying zoonotic arbo- diverse mosquito and vertebrate communities. viruses in Australia and provides a framework for other ar- Whether a dilution effect exists is an important question boviruses such as urban flaviviruses, which may be reliant on when dealing with zoonotic pathogens with multiple reser- urban bird populations for transmission (Maute et al., 2019). voir hosts (Rohr et al., 2019). The dilution effect applies to The nature of the data collection, both for mosquito and situations in which species diversity reduces measures of vertebrate communities, is time consuming and costly. We disease risk because not all hosts are susceptible and, there- found that the use of a single mosquito trap in each SSC fore, act as decoys for vector-borne pathogens (Keesing et al., successfully captured a diversity of mosquito communities 2006). This study was not explicitly designed to detect a that largely corresponded to the habitat conditions (for 10 SKINNER ET AL.
example, frequent trapping of Ae. vigilax at Bracken Ridge, a tion with human notification rates of RRV in selected suburbs saltwater habitat; or frequent trapping of Cx. annulirostris at of Brisbane. Further investigations, including ongoing ver- Corinda and Indooroopilly, freshwater habitats). Future tebrate surveillance and modeling studies, are needed to studies, however, would benefit from additional traps. There quantify the importance of vector competence and mosquito would be great value for future investigations to expand data feeding patterns in association with the novel finding of collection with additional SSCs across the gradient of human vertebrate biomass from this study. notifications, as well as additional mosquito traps and host surveys. Acknowledgments A caveat of selecting study sites on the basis of human no- tifications is that it may not accurately represent locations with The collection of these data would not have been possible high or low infection rates of RRV in humans. First, the place of without the following field volunteers: Georgia Braun, Cara residence of an infected person may not have been the place of Parsons, Ian Parsons, Hannah Thomas, Jack Dodd, Scout infection. In addition, not all cases of RRV cause clinical Fisher, Stevie Tozer, Stuart MacLeod, Robin Rowland, Jay- manifestations, and the proportion of clinical cases that result in lan Schabrod, Michael Johnson, and Trina Kateifides. The notifications is unknown. Finally, it is unknown whether authors acknowledge Lara Herrero for supervising the PhD asymptomatic cases in humans contribute to RRV transmis- study of which this research is a part and more broadly HM sion. These factors could result in an underestimate in RRV laboratory members and Lara Herrero laboratory members transmission and notified cases in humans. With these un- for feedback on data collection and analysis. certainties in mind, and given the small number of annual no- tifications within each SSC (n = 0–26), we determined that a Author Disclosure Statement long-term notification rate is most suitable to identify and cat- No conflicting financial interests exist. egorize those areas with tendencies for high or low RRV cases. Future studies on RRV community ecology would benefit Funding Information from including broader vertebrate survey methods (such as active trapping or passive camera trapping) to incorporate a This project was supported by an Ecological Society of wider diversity of vertebrates, including day-active species. Australia Graduate Grant. E.B.S. was supported by an Aus- For example, small and/or cryptic vertebrates, such as mu- tralian Government Research Training Program and a Grif- rids, were infrequently detected in this study. However, some fith Graduate Research School Publication Assistance serological evidence suggests that these species are exposed Scholarship. A.J.P. was supported by a Queensland Gov- to RRV (Vale et al., 1991). Although limited in their appli- ernment Accelerate Postdoctoral Research Fellowship. cation, open-access biodiversity datasets (such as the Global Biodiversity Information Facility [Flemons et al., 2007] and Supplementary Material the Atlas of Living Australia [Belbin and Williams, 2016]) Supplementary Figure S1 could be used to cross-check species lists or identify verte- Supplementary Table S1 brates in the absence of formal vertebrate surveys. Supplementary Table S2 In addition, inclusion of mosquito bloodmeal analyses Supplementary Table S3 would further inform the potential role of the vertebrates Supplementary Table S4 considered in this study, by demonstrating which mosquito vectors readily interact with which vertebrate hosts. Of three published blood-feeding studies conducted in Brisbane be- References tween 1995 and 2008, birds, humans, placental mammals, and Aaskov JG, Mataika JU, Lawrence GW, Rabukawaqa V, et al. marsupials were all variously dominant among the vertebrate An epidemic of Ross River virus infection in Fiji. Am J Trop species fed on by mosquitoes, but this was likely influenced by Med Hyg 1981; 30:1053–1059. differing mosquito collection sites between the studies (Ryan Australian Bureau of Statistics. Australian Standard Geo- et al., 1997, Kay et al., 2007, Jansen et al., 2009). Only one of graphical Classification (AGSC). 1999. Available at https:// www.abs.gov.au/ausstats/[email protected]/Lookup/by%20Subject/
Downloaded by Queensland University of Technology from www.liebertpub.com at 08/17/20. For personal use only. these three studies attempted to quantify the relative avail- ability of host species in trapping sites, finding that, overall, 1270.0.55.001~July%202016~Main%20Features~Non%20 dogs were the most dominant bloodmeal relative to their ABS%20structures *10008 abundance, but this was based largely on subjective house- Australian Bureau of Statistics. 2016 Census QuickStats. 2016. holder survey (rather than vertebrate observation surveys) and Available at https://quickstats.censusdata.abs.gov.au/census_ was undertaken in SSCs where animals with higher biomass services/getproduct/census/2016/quickstat/3GBRI?opendocument Australian Government. National Notifiable Diseases Surveil- such as horses were not abundant (Kay et al., 2007). This also lance System. 2018. Available at http://www9.health.gov.au/ supports the inclusion of vertebrate and bloodmeal surveys cda/source/cda-index.cfm plus consideration of the effect of biomass in future studies. Australian Government. Ross River Virus Infection Case De- finition. Department of Health. 2019. Available at https:// Conclusion www1.health.gov.au/internet/main/publishing.nsf/Content/cda- surveil-nndss-casedefs-cd_rrv.htm Improved understanding of both mosquito and vertebrate Azuolas JK. Ross River virus disease of horses. Aust Equine communities can lead to better prediction and management of Vet 1998; 16:56–58. transmission for zoonotic mosquito-borne diseases. In this Belbin L, Williams KJ. Towards a national bio-environmental study, we find that RRV vertebrate biomass (particularly data facility: Experiences from the Atlas of Living Australia. horses) and mosquito abundance had the strongest associa- Int J Geogr Inf Sci 2016; 30:108–125. VECTOR–HOST COMMUNITY ECOLOGY & HUMAN INFECTION 11
Boyd AM, Hall RA, Gemmell RT, Kay BH. Experimental in- lecular analysis of avian blood meals. Am J Trop Med Hyg fection of Australian brushtail possums, Trichosurus vulpe- 2009; 81:849–857. cula (Phalangeridae: Marsupialia), with Ross River and Jardine A, Neville PJ, Lindsay M. Proximity to mosquito breeding Barmah Forest viruses by use of a natural mosquito vector habitat and Ross River virus risk in the peel region of Western system. Am J Trop Med Hyg 2001; 65:777–782. Australia. Vector-Borne Zoonotic Dis 2015; 15:141–146. Boyle TP, Smillie GM, Anderson JC, Beeson DR. A sensitivity Jeffery JA, Ryan PA, Lyons SA, Kay BH. Vector competence analysis of nine diversity and seven similarity indices. Res J of Coquillettidia linealis (Skuse) (Diptera: Culicidae) for Water Pollut Control Fed 1990; 749–762. Ross River and Barmah Forest viruses. Aust J Entomol 2002; Chapman HF, Hughes JM, Jennings C, Kay BH, et al. Popu- 41:339–344. lation structure and dispersal of the saltmarsh mosquito Aedes Jones KE, Bielby J, Cardillo M, Fritz SA, et al. PanTHERIA: A vigilax in Queensland, Australia. Med Vet Entomol 1999; 13: species-level database of life history, ecology, and geography 423–430. of extant and recently extinct mammals. Ecology 2009; 90: Claflin SB, Webb CE. Ross River virus: Many vectors and 2648. unusual hosts make for an unpredictable pathogen. Plos Pa- Kay BH, Boyd AM, Ryan PA, Hall RA. Mosquito feeding thog 2015; 11:e1005070. patterns and natural infection of vertebrates with Ross River Crocker W, Maute K, Webb C, French K. Mosquito assem- and Barmah Forest viruses in Brisbane, Australia. Am J Trop blages associated with urban water bodies; implications for Med Hyg 2007; 76:417–423. pest and public health threats. Landscape Urban Plan 2017; Kay BH, Hall RA, Fanning ID, Mottram P, et al. Experimental 162:115–125. infection of vertebrates with Murray Valley encephalitis and Eyre TJ, Ferguson DJ, Hourigan CL, Smith GC, et al. Terres- Ross River viruses. Paper presented at the Arbovirus Re- trial Vertebrate Fauna Survey Assessment Guidelines for search in Australia, 1986. Queensland. Brisbane: Department of Science. Information Kay BH, Pollitt CC, Fanning ID, Hall RA. The experimental Technology, Innovation and the Arts, Queensland Govern- infection of horses with Murray Valley encephalitis and Ross ment, 2014. River viruses. Aust Vet J 1987; 64:52–55. Flemons P, Guralnick R, Krieger J, Ranipeta A, et al. A web- Keesing F, Holt RD, Ostfeld RS. Effects of species diversity on based GIS tool for exploring the world’s biodiversity: The disease risk. Ecol Lett 2006; 9:485–498. Global Biodiversity Information Facility Mapping and Ana- Kilpatrick AM, Daszak P, Jones MJ, Marra PP, et al. Host lysis Portal Application (GBIF-MAPA). Ecol Inform 2007; 2: heterogeneity dominates West Nile virus transmission. Proc R 49–60. Soc Lond B Biol Sci 2006; 273:2327–2333. Franz R, Soliva CR, Kreuzer M, Steuer P, et al. Methane pro- Kindt R. Package ‘BiodiversityR’. 2019. Available at http://vps. duction in relation to body mass of ruminants and equids. fmvz. usp. br/CRAN/web/packages/BiodiversityR Evol Ecol Res 2010; 12:727–738. Knight F, Menkhorst P. Field Guide to Mammals of Australia, Gardner AM, Muturi EJ, Overmier LD, Allan BF. Large-scale Melbourne: Oxford University Press, 2010. removal of invasive honeysuckle decreases mosquito and Lee DJ, Hicks MM, Griffiths M, Russell RC, et al. The Culi- avian host abundance. Ecohealth 2017; 14:750–761. cidae of the Australian Region. Canberra: Australian Gov- Garnett ST, Duursma DE, Ehmke G, Guay P-J, et al. Biological, ernment Publishing Service, 1989:7. ecological, conservation and legal information for all species Lindsay M, Oliveira N, Jasinska E, Johansen C, et al. An out- and subspecies of Australian bird. Sci Data 2015; 2:150061. break of ross river virus disease in southwestern Australia. Gummow B, Tan RHH, Joice RK, Burgess G, et al. Ser- Emerg Infect Dis 1996; 2:117–120. oprevalence and associated risk factors of mosquito-borne Lindsay M, Oliveira N, Jasinska E, Johansen C, et al. Western alphaviruses in horses in northern Queensland. Aust Vet J Australia’ s largest recorded outbreak of Ross River virus 2018; 96:243–251. disease. Arbov Res Aust 1997; 7:147–152. Hanford JK, Webb CE, Hochuli DF. Habitat traits associated Lyimo IN, Ferguson HM. Ecological and evolutionary deter- with mosquito risk and aquatic diversity in urban wetlands. minants of host species choice in mosquito vectors. Trends Wetlands 2019; 39:743–758. Parasitol 2009; 25:189–196. Harley D, Sleigh A, Ritchie S. Ross River virus transmission, Maute K, Webb C, Phalen D, Martin J, et al. Clean bill of infection, and disease: A cross-disciplinary review. Clin Mi- health? Towards an understanding of health risks posed by crobiol Rev 2001; 14:909–932. urban ibis. J Urban Ecol 2019; 5:juz006. Downloaded by Queensland University of Technology from www.liebertpub.com at 08/17/20. For personal use only. Hill CA, Kafatos FC, Stansfield SK, Collins FH. Arthropod- Morcombe MK. Field Guide to Australian Birds. Brisbane: borne diseases: Vector control in the genomics era. Nat Rev Steve Parish Publishing, 2003. Microbiol 2005; 3:262. Oksanen J, Blanchet GF, Kindt R, Legendre P, et al. Vegan: Hoffmann AA, Ross PA, Rasˇic´ G. Wolbachia strains for disease Community ecology package. R package version 1.17–4. control: Ecological and evolutionary considerations. Evol 2010. http://cran. r-project. org> Appl 2015; 8:751–768. Queensland Government. Notifiable conditions annual reporting. Janousek WM, Marra PP, Kilpatrick AM. Avian roosting be- 2019. Available at https://www.health.qld.gov.au/clinical- havior influences vector-host interactions for West Nile virus practice/guidelines-procedures/diseases-infection/surveillance/ hosts. Parasit Vectors 2014; 7:399. reports/notifiable/annual Jansen CC, Shivas M, May F, Pyke A, et al. Epidemiologic, Roberts BJ, Catterall CP, Eby P, Kanowski J. Long-distance entomologic, and virologic factors of the 2014–15 Ross River and frequent movements of the flying-fox Pteropus polioce- virus outbreak, Queensland, Australia. Emerg Infect Dis phalus: Implications for management. PLoS One 2012; 7: 2019; 25:2243–2252. e42532. Jansen CC, Webb CE, Graham GC, Craig SB, et al. Blood Rohr JR, Civitello DJ, Halliday FW, Hudson PJ, et al. Towards sources of mosquitoes collected from urban and peri-urban common ground in the biodiversity–disease debate. Nat Ecol environments in eastern Australia with species-specific mo- Evol 2019; 4:24–33. 12 SKINNER ET AL.
Russell RC. Constructed wetlands and mosquitoes: Health Takken W, Verhulst NO. Host preferences of blood-feeding hazards and management options—An Australian perspec- mosquitoes. Ann Rev Entomol 2013; 58:433–453. tive. Ecol Eng 1999; 12:107–124. Tidemann CR, Nelson JE. Long-distance movements of the Russell RC. Ross River virus: Ecology and distribution. Ann grey-headed flying fox (Pteropus poliocephalus). J Zool Rev Entomol 2002; 47:1–31. 2004; 263:141–146. Ryan PA, Do KA, Kay BH. Definition of Ross River virus Vale TG, Spratt DM, Cloonan MJ. Serological evidence of vectors at Maroochy Shire, Australia. J Med Entomol 2000; arbovirus infection in native and domesticated mammals on 37:146–152. the south coast of New South Wales Australia. Aust J Zool Ryan PA, Do KA, Kay BH. Spatial and temporal analysis of 1991; 39:1–8. Ross River virus disease patterns at Maroochy Shire, Aus- van den Hurk AF, Craig SB, Tulsiani SM, Jansen CC. Emerging tralia: Association between human morbidity and mosquito tropical diseases in Australia. Part 4. Mosquito-borne dis- (Diptera: Culicidae) abundance. J Med Entomol 1999; 36: eases. Ann Trop Med Parasitol 2010; 104:623–640. 515–521. Webb CE, Russell RC. Dispersal of the mosquito Aedes vigilax Ryan PA, Martin L, Mackenzie JS, Kay BH. Investigation of (Diptera: Culicidae) from urban estuarine wetlands in Syd- gray-headed flying foxes (Pteropus poliocephalus) (Mega- ney, Australia. J Med Entomol 2019; 56:1290–1295. chiroptera: Pteropodidae) and mosquitoes in the ecology of Webb CE, Russell R, Doggett S. A Guide to Mosquitoes of Ross river virus in Australia. Am J Trop Med Hyg 1997; 57: Australia. Australia, Csiro Publishing, 2016. 476–482. Weinstein P. An ecological approach to public health inter- Smith P, Smith J. Climate change and bird migration in south- vention: Ross River virus in Australia. Environ Health Per- eastern Australia. EMU-Aust Ornithol 2012; 112:333–342. spect 1997; 105:364–366. Stephenson EB, Murphy AK, Jansen CC, Peel AJ, et al. Inter- preting mosquito feeding patterns in Australia through an Address correspondence to: ecological lens: An analysis of blood meal studies. Parasit Eloise B. Skinner Vectors 2019b; 12:156. Environmental Futures Research Institute Stephenson EB, Peel AJ, Reid SA, Jansen CC, et al. The non- Griffith University human reservoirs of Ross River virus: A systematic review of Nathan 4111 the evidence. Parasit Vectors 2018; 11:188. Queensland Stephenson EB. Ross River virus seroprevalence and associated Australia life-history traits in non-human vertebrates across South-East Queensland. (2019a) Doctoral Thesis, Griffith University. E-mail: eloise.stephenson@griffithuni.edu.au Downloaded by Queensland University of Technology from www.liebertpub.com at 08/17/20. For personal use only.