Cautioning the use of degree-day models for climate change projections in the presence of parametric uncertainty

Authors

  • Julia L. Moore,

    1. Program in Population Biology, Ecology and Evolution, Graduate Division of Biological and Biomedical Sciences, Emory University, 1510 Clifton Rd, Atlanta, Georgia 30322 USA
    Search for more papers by this author
    • Present address: Department of Evolution and Ecology, Population Biology Graduate Group, One Shields Avenue, University of California, Davis, California 95616 USA.

  • Song Liang,

    1. Department of Environmental and Global Health, College of Public Health and Health Professions, and Emerging Pathogens Institute, University of Florida, P.O. Box 100188, Gainesville, Florida 32610 USA
    Search for more papers by this author
  • Adam Akullian,

    1. Department of Epidemiology, School of Public Health and Community Medicine, University of Washington, Box 357236, Seattle, Washington 98195 USA
    Search for more papers by this author
  • Justin V. Remais

    Corresponding author
    1. Program in Population Biology, Ecology and Evolution, Graduate Division of Biological and Biomedical Sciences, Emory University, 1510 Clifton Rd, Atlanta, Georgia 30322 USA
    2. Department of Environmental Health, Rollins School of Public Health, Emory University, 1518 Clifton Rd Room 2023, Atlanta, Georgia 30322 USA
    Search for more papers by this author

  • Corresponding Editor: M. P. Ayres.

Abstract

Developmental models, such as degree-day models, are commonly used to predict the impact of future climate change on the intensity, distribution, and timing of the transmission of infectious diseases, particularly those caused by pathogens carried by vectors or intermediate hosts. Resulting projections can be useful in policy discussions concerning regional or national responses to future distributions of important infectious diseases. Although the simplicity of degree-day models is appealing, little work has been done to analyze their ability to make reliable projections of the distribution of important pathogens, vectors, or intermediate hosts in the presence of the often considerable parametric uncertainty common to such models. Here, a population model of Oncomelania hupensis, the intermediate host of Schistosoma japonicum, was used to investigate the sensitivity of host range predictions in Sichuan Province, China, to uncertainty in two key degree-day model parameters: δmin (minimum temperature threshold for development) and K (total degree-days required for completion of snail development). The intent was to examine the consequences of parametric uncertainty in a plausible biological model, rather than to generate the definitive model. Results indicate that model output, the seasonality of population dynamics, and range predictions, particularly along the edge of the range, are highly sensitive to changes in model parameters, even at levels of parametric uncertainty common to such applications. Caution should be used when interpreting the results of degree-day models used to generate predictions of disease distribution and risk under scenarios of future climate change, and predictions should be considered most reliable when the temperature ranges used in projections resemble those used to estimate model parameters. Given the potential for substantial changes in degree-day model output with modest changes in parameter values, caution is warranted when results will be used to inform policy and management decisions.

Ancillary