November 1, 2017

Analysis of the effects of climate and environment on Fritillaria gentneri flowering

Erin C. Gray, Matt A. Bahm, and Meaghan I. Petix | 2017

  We used non-parametric multiplicative regression (NPMR) to model the effects of climate and environmental variables on flowering of Fritillaria gentneri from 57 sites throughout its range in southern Oregon.

 

  • We used non-parametric multiplicative regression (NPMR) to model the effects of climate and environmental variables on flowering of Fritillaria gentneri from 57 sites throughout its range in southern Oregon. We report 4 models using different sites in an attempt to find the strongest predictors for flowering of F. gentneri:
    • In Model 1(51 sites), number of flowering F. gentneri was best explained by the previous winter’s precipitation (in), and the previous spring’s minimum temperature (F). While this model had a strong p value, these variables were able to capture only 7.5% of the variability in the data.
    • For Model 2, sites with a mean of less than one were deleted, resulting in 36 sites. In this model, spring precipitation (in) and the previous spring’s minimum temperature (F) were the strongest predictors, explaining 9.7 % of the variability, which was a slight improvement from Model 1.
    • For Model 3, we deleted sites that had zero plants in 5 or more years, resulting in 32 sites. Number of flowering plants was best explained by spring precipitation (in), previous spring’s minimum temperature (F), and previous winter maximum vapor pressure deficit.  Collectively these predictors explained 15% of the variability in the data, resulting in the strongest model.
    • For Model 4, we used a more long-term data set of 24 sites monitored from 1999-2017. Number of flowering plants was best explained by previous winter maximum vapor pressure deficit and winter maximum vapor pressure deficit; these predictors only explained 5.8% of the response.
  • While our best models only explained up to 15% of the variability in the data, they do suggest climate variables that were consistently chosen as predictors are likely to impact flowering of F. gentneri across its range.