Quantifying the Effects of Multiple Environmental Factors on Plant Demography: Prediction and Inference in a Multivariate World

All populations are affected simultaneously by multiple environmental factors, yet most plant demography studies examine single factors in isolation. Failing to account for multiple environmental factors can lead to inaccurate inferences and predictions. We identify four key complications associated with inference and prediction in multifactor studies: ( a ) correlations among factors, ( b ) nonadditive effects of factors on vital rates, ( c ) nonadditivity arising from the effects of factors on the sensitivity of population growth rate to vital rates, and ( d ) violations of space-for-time substitution. We review the plant population ecology literature to assess how frequently multiple environmental factors are considered and how rigorously multifactor studies address each complication. We find that most studies address only one factor, and multifactor studies address few of these challenges. We synthesize conceptual and empirical work addressing each complication and outline practical recommendations for improving inference and prediction in multivariate demographic studies.