Event Abstract Back to Event Avian Influenza in California: Does El Niño Southern Oscillation affect disease presence? Narjara V. Grossmann1*, Kathryn Conlon2 and Beatriz Martínez-López1 1 Center for Animal Disease Modeling and Surveillance, School of Veterinary Medicine, University of California, Davis, United States 2 Department of Medicine & Epidemiology, School of Veterinary Medicine, University of California, Davis, United States High risk areas and important environmental factors contributing to avian influenza in multiple regions1,2,3 as well as in California4 have been detected using Ecological niche models. The surge of highly pathogenic avian influenza in domestic poultry has been also proved to be related to the presence and contact with migratory birds4. However, few studies have evaluated the effect of other seasonal and climatic changes, such as those that accompany the El Niño phenomenon. El Niño Southern Oscillation (ENSO) is a climatic phenomenon characterized by two phases: an increase in ocean temperatures (El Niño) and a decrease (La Niña). The NOAA (National Oceanic and Atmospheric Administration) monitors this phenomenon by reporting the average three-month sea surface temperature (https://www.climate.gov/enso). This event is accompanied by global changes in precipitation and wind patterns, which indirectly may affect distribution patterns and behavior of migratory birds. For the state of California, there are many evidences that El Niño promotes an increase in precipitation, in terms of quantity and spatial distribution6,7. During the 1992-93 cycle, rainfall increased over 5 times the historical mean for arid islands in the Gulf of California and the biological consequences included a 160-fold increase in plant cover and a doubling of insect abundance8. Coastal productivity follows a similar pattern as those observed for the Peruvian/Equatorial coast9. Migratory birds can be affected by these events especially when migrating from areas that have reduced precipitation caused by El Niño. During La Niña, the inverse relationship is expected, that is, a decrease in precipitation affecting the vegetation cover, insect and rodent population and subsequently their avian consumers10,11. We believe that avian influenza cases among wild bird populations are influenced by seasonal changes in climate such as ENSO phenomenon. To evaluate this we extracted raw information from the Influenza Research Database (https://www.fludb.org). Month of the year and season were extracted from this dataset and added as new columns, as well as the tri-monthly sea surface temperature (https://www.climate.gov/enso). Temperature, snow cover and precipitation were extracted from the NOAA database usng the rnoaa package though the R interface. So far, preliminary analyses indicated that during El Niño years there are overall fewer cases of avian influenza compared to La Niña. This detection is apparently not associated with differences in sampling efforts. Normal years do not seem to follow a pattern, therefore other factors can be associated to predicting the high-risk season for occurrence of avian influenza. To further evaluate this relationship we looked at the years, 2006 (a normal year in terms of ENSO), 2009 (El Niño year) and 2010 (strong La Nina). A geographic weighted regression was done to all three datasets that consisted of: case ratio (dependent variable), temperature, precipitation and snowfall. For 2006 precipitation did not seem to be strongly associated to positive cases, whereas temperature was more important. For both 2010 and 2009 temperature was no longer important, and precipitation was the best predicting variable. However, these preliminary results should be evaluated with caution since they only represent one year for each type of ENSO variation. Also, these models explained around 20 - 24% of the observed variation, therefore, other variables might be more important in predicting cases than those that were included in the model. These values indicate that we are in the right direction and should gather more data to create more robust models. Considering that this year is an El Nino year, and that 2020 is predicted to be a strong La Niña year, if our predictions are correct, we might see a surge of cases in the coming year. A model that can account for that may better prepare us for these events. Acknowledgements This project was conducted in part thanks to the UC Davis Wildlife Health Center fellowship.
Read full abstract