Effects of seasonal environments on stage-structured, rare plant populations

Background/Question/Methods

Predicting the effects of climate change on plant species requires an understanding of the relationship between population vital rates and environmental drivers. Fluctuations in populations through time are attributed to stochastic environments, and therefore usually simulated under independent and identically distributed scenarios, as in demographic models. This parsimonious assumption disregards biotic and abiotic temporal relationships with population growth at the cost of large prediction error. Previous studies explicitly modeled environmental disturbances on plant populations to increase model reliability and estimates of population risk, but studies of how plant population vital rates are affected by seasonal environmental effects are rare due to a lack of long-term data on a single population. This study assesses the relationship of five abiotic drivers with population growth rates using nine years of demographic data from Astragalus tyghensis, a rare, native, Oregon perennial. We used non-metric multidimensional scaling, typically used for communities, to determine the strongest abiotic correlate with stage-structured population vital rates for each species. Our objectives were to find a potential environmental driver of a multivariate, population response, and to assess the type of relationship with each vital rate.

Results/Conclusions

Preliminary results show a moderately strong relationship (R² = 32.8%) between mean dry growing season evapotranspiration rates and a two-dimensional ordination, with variation explained along a growth and fecundity axis, for five, proximal populations of Astragalus tyghensis. Evapotranspiration was positively correlated with growth and fecundity rates in the populations and negatively correlated with retrogression to smaller stages. Extremely dry years are expected to negatively effect population growth, but we observed a positive effect, suggesting an indirect effect such as reduction in competition. NMS however does not account for interacting or non-linear relationships among factors, which are likely behind the mechanisms involved in regulating population fluctuations. Still, this is a strong improvement in model sophistication and assumptions over previous approaches, with emphasis on biological realism that is necessary to project population risk given severe, forecasted changes in climate. Further, to help mitigate the loss of threatened and endangered populations, the broad study implications include the assisted migration of at-risk populations into suitable habitats based on environmental driver effects and climate forecasts.