Print PageClimate change, such as increased temperature and decreased precipitation, could reduce wetland availability as much as 47%. Since breeding success of most duck species, as well as within-season movement of many duck species, are heavily dependent on wetlands conditions, climate change will have a significant effect on duck breeding populations. In this study, we obtain duck population data from Waterfowl Breeding Population and Habitat Survey and corresponding weather data from Environment Canada, and use Singular Value Decomposition (SVD) to fill the missing weather data. We then use Partial Least Square Regression (PLSR) to find the relation between weather and duck population.
Results/Conclusions
Our results show that SVD is reliable method to fill missing data by combining both temporal and spacial trends. For PLSR, spring temperature could explain about 60% variation in duck population, which is 50% for winter temperature, 60% for yearly precipitation, and 90% if all weather information is included; both spring temperature and winter temperature are useful to increase the predictive ability of the regression model, which is consistent with previous studies that wetlands are more sensitive to temperature than to precipitation.
