Projects per year
Abstract
Millions of birds undertake seasonal migration between breeding and wintering sites. Bird migration causes various ecological effects such as affecting local predator-prey relationships and transporting pathogens, seeds and energy. Among these effects, pathogen dispersal has caused a large debate, including how migratory birds disperse pathogens, and how migratory birds interact with pathogens during their migration. A better understanding of pathogen dispersal is urgently needed because it is relevant to both wildlife and human health. Therefore, empirical studies such as spatial-temporal correlations between infection outbreaks and migration trajectories, genetic studies between outbreaks and infection dynamics in migratory populations, and theoretical modelling have been carried out.
Although previous studies suggested that bird migration can disperse pathogens along migration route of birds, however, bird migration may also reduce infection prevalence and limit pathogen dispersal by so-called ‘migratory escape’ and ‘migratory culling’. Therefore, migration can affect pathogen dispersal and infection prevalence in a population, but its effects may vary among host-pathogen systems.
Most migratory bird species use stopover sites where they refuel and rest during their migration. The movement of birds connects these stopover sites, together with their breeding and wintering sites, in a migration network. Some stopover sites are selected over others, and this selection varies between species, and over time and space within a certain species, so configurations of migration network change, and can be characterised by ‘serial stopover sites’ (when the birds are migrating over a narrow front) or ‘parallel stopover sites’ (when migration occurs over a broad front). These patterns have been clearly observed by previous studies through satellite telemetry tracking.
Apart from various spatial configurations, migratory birds vary their timing of departure as well, and this synchrony in departure can vary from weeks to months. The combinations of the various patterns in network configuration and departure synchrony influence aggregation size, resting duration at stopover sites, and contact probabilities among individuals. However, the effects of network configuration and migration synchrony on pathogen dispersal and infection dynamic has not been fully examined yet.
Furthermore, although stopover sites are crucial for migratory birds to complete their migration, the availability of suitable stopover sites in the East Asian-Australasian Flyway decreased, especially in China, where 30% of natural wetlands were lost over the last two decades. In reaction to this wetland loss, the migration network becomes restricted to fewer remaining sites, and bird abundance on the remaining sites correspondingly increases. This intensive use of remaining sites may increase the probability of site infection and infection prevalence in the population. However, the impact of wetland loss on infection dynamics has not been investigated before.
Network analysis is a promising tool to analyse pathogen dispersal by migratory birds. For example, it was used to study the dispersal of severe acute respiratory syndrome, and foot and mouth disease. Real world networks such as trade networks and transport networks, are often recognized as scale-free networks. Such networks are very efficient in dispersing pathogens over the network. When extensive habitat loss occurs, however, the scale-free topology could disappear, which can make pathogen dispersal among sites less effective. Alternatively, the infection prevalence in migratory birds might be increased due to larger aggregations at remaining sites. However, the topologies of bird migration networks have rarely been examined in empirical studies or in theoretical work, although it could provide a better understanding of the variables that influence pathogen dispersal.
In this study, I focused on avian influenza viruses (AIVs), an influenza virus that is adapted to infect birds, especially waterfowl, such as many duck, goose and swan species. It infects mainly birds, but in some cases, it can also infect mammals such as swine, horses, whales, bats and humans. Avian influenza viruses can be classified into two groups: low pathogenic avian influenza viruses (LPAIVs) and highly pathogenic avian influenza viruses (HPAIVs), based on the severity of the illness that they cause in chickens. Infection of LPAIVs in wild birds only causes mild symptoms, however, subtypes H5 and H7 can mutate to HPAIV when multiple low pathogenic avian influenza subtypes co-infect one host, especially in poultry farms with low bio-security and large numbers of domestic birds.
HPAIV attracted a lot of attention due to their rapid dispersal and large impacts. For example, the highly pathogenic avian influenza H5N1 was for the first time detected in a domestic goose in 1996 in Guangdong, China, and then detected in wild birds in 2002 in Hong Kong, China. It suggested that migratory wild birds were infected with avian influenza virus from domestic birds. Furthermore, a H5N1 outbreak was detected in wild birds in 2005 at Qinghai Lake, China, killing more than 6000 birds. Within a few months, the H5N1 was detected in Europe, Middle East and Africa. Although intensive studies have focused on dispersal of avian influenza virus, most of these were carried out in duck species such as mallard. Few studies have examined the role of other migratory waterfowl, such as goose species. Since HPAIV can spill-over to humans and could cause high mortality rates, it is urgent to understand the mechanism of avian influenza virus dispersal.
Overall, the aim of this study is to obtain a better understanding of the impact of migration on dispersing avian influenza virus by combining modelling and spatial-temporal statistical approaches.
In chapter 2, I examined the infection dynamic of LPAIV in migratory goose species. I analysed throat and cloaca samples that were collected from three species from their breeding sites, stopover sites and wintering sites. I examined the infection prevalence on these sites, and analysed the temporal patterns in infection prevalence. My results showed that migratory geese were probably not infected with LPAIVs before arrival on their wintering sites, as they had a relatively low infection prevalence just after the arrival, but the prevalence increased over the winter period. My results suggest that migratory geese were exposed to the LPAIV shortly after their arrival, indicating that they might not disperse the virus during autumn migration, but more likely disperse it during spring migration.
In chapter 3, I examined the effects of spatial and temporal migration patterns on the dynamics of low pathogenic avian influenza infection prevalence. I applied a discrete-time SIR (Susceptible-Infected-Recovered) model, with environmental transmission and migration, to various migration strategies, including networks with serial, and/or parallel stopover sites, and with various levels of migration synchrony. My results showed that both an increase in the number of serial stopover sites and an increase in the synchrony of departure timing reduces the infection prevalence due to ‘migratory escape’. Whereas increasing the number of parallel stopover sites increases the infection prevalence, because the migratory population is exposed to a larger total amount of virus in the environment, speeding-up the accumulation of infections. Furthermore, my simulations suggest that if migratory species adopt a migration pattern with multiple serial stopover sites and with high migration synchrony, the AIV transmission becomes less efficient in the population, and thereby lead to a low infection prevalence.
In chapter 4, I tested whether habitat loss facilitates pathogen dispersal and infection prevalence in a migratory population. I identified all potential stopover sites of greater white-fronted geese in the East Asian-Australasian Flyway, and constructed migration networks with various habitat loss scenarios. I used Agent-based models to simulate bird migration over various migration networks, and integrated these with SIR-type infection dynamics to simulate epidemiological processes. I studied the dynamic of infection prevalence in migratory populations and the infection probability at stopover sites under various habitat loss scenarios. Consistent with my previous findings, I found that migration can reduce infection prevalence in a population due to migratory escape. However, the population cannot lose infection completely due to a relay effect that resting birds can be infected with avian influenza viruses that were shed by previous resting birds. Moreover, under severe levels of habitat loss, i.e., removing all sites with area decrease, geese start aggregating earlier in the fewer remaining sites, and thereby facilitate infection. In addition, habitat loss increases the infection probability for the remaining sites due to a larger amount of visiting birds, which potentially carry the virus. These results thus suggest that habitat loss facilitates outbreak of avian influenza virus infection in a migratory population and increases the probability for pathogen dispersal.
In chapter 5, I summarized the historical HPAIV outbreaks in swan goose and bar-headed goose and compared their contact opportunities with avian influenza outbreaks areas and with the distribution areas of domestic birds in their migration corridors. Their migration corridors were estimated from GPS tracking data, using a dynamic Brownian Bridge Movement Models (dBBMMs). I found that swan geese were more likely to come into contact with outbreak areas, but fewer outbreaks occurred in their population. In contrast to swan geese, bar-headed geese were less likely to come into contact with outbreak areas, but more outbreaks occurred in their population. Moreover, I found that the densities of domestic ducks in the migratory corridor of swan geese were higher compared with those of bar-headed geese. On the basis of these findings, I proposed two possible explanations for these contrast infection patterns. First, frequent contact and a long contact history with domestic ducks may have caused higher levels of innate immunity in swan goose. Second, the migration strategy of bar-headed goose may compromise immunity, so that bar-headed geese are more vulnerable to HPAIV.
These studies broaden the knowledge of bird species’ roles in affecting avian influenza virus infection dynamic and the virus dispersing during seasonal migration. The environmental transmission plays an important role in keeping the virus circulating in a migratory population, and I therefore recommend increasing the efforts for monitoring virus concentrations in water bodies used during migration. Moreover, since swan goose may have higher levels of innate immunity and be more resistant to infection by HPAIVs, I also recommend increasing active surveillance that covers not only the well-known affected goose species such as bar-headed goose, but also goose species which are less often found dead due to infection with HPAIVs.
Although this study focused on the interactions between host migration and infectious pathogens in the goose-AIVs system, the findings can be generalized to other migratory host-pathogen systems such as butterflies-parasite systems, if the pathogen can persist in the environment.
Original language | English |
---|---|
Qualification | Doctor of Philosophy |
Awarding Institution |
|
Supervisors/Advisors |
|
Award date | 4 Dec 2018 |
Place of Publication | Wageningen |
Publisher | |
Print ISBNs | 9789463435048 |
DOIs | |
Publication status | Published - 4 Dec 2018 |
Fingerprint
Dive into the research topics of 'Consequences of seasonal migration: How goose relocation strategies influence infection prevalence and pathogen dispersal'. Together they form a unique fingerprint.Projects
- 1 Finished
-
Impact of migration on prevalence of avian influenza infection in a migratory populatio: a modelling study.
Yin, S. (PI), Prins, H. (CoI), de Boer, F. (CoI), Yin, S. (PhD candidate), Prins, H. (Promotor) & de Boer, F. (Promotor)
1/09/14 → 4/12/18
Project: PhD