Abstract Detail



Biodiversity Informatics & Herbarium Digitization

Love, Natalie [1], Park, Isaac [2], Mazer, Susan [2].

A new phenological metric for use in pheno-climatic models: a case study using herbarium specimens of Streptanthus tortuosus.

Herbarium specimens provide an extensive geographic and temporal record of phenology and can be used to detect temporal shifts in flowering time and sensitivities to climate. Previous studies that have investigated these relationships have used the day of year of collection (DOY) of herbarium specimens as a proxy for either first flowering date or date of peak flowering.  The high variation among herbarium sheets with respect to the phenological status of the sampled plants, however, undermines the assumption that the DOY represents any particular phenological phase.  Ignoring this variation can reduce the explanatory power of models that aim to predict the effects of climatic conditions on the date on which any particular phenophase appears or is observed. Here, we develop a new protocol for scoring phenology on digitized herbarium specimens using a free ImageJ plugin, Cell Counter; we use these counts to present a new quantitative metric of the phenological status of herbarium specimens – the phenological index (PI). We then use this index in pheno-climatic models to control for phenological variation among sampled plants when testing for the effects of climatic conditions on the DOY; and we illustrate how the inclusion of the PI as an independent variable can improve the predictive capacity of pheno-climatic models. Including the PI in multivariate pheno-climatic models increased the R2by 31% and reduced model error. We also measured the phenological sensitivity of the mountain jewelflower [Streptanthus tortuosus (Brassicaceae)] to variation in local spring Tmax and winter rainfall. Our phenological scoring protocol provides a simple way to score phenology from imaged herbarium specimens and is readily available to all researchers. Including PI in pheno-climatic models improves the predictive power of such models and allows these models to be used to forecast the DOY of any phenophase (e.g., peak flower) at a given location and climate. 


1 - 296 N Hope Ave., SPC 25, Santa Barbara, CA, 93110, United States
2 - University of California, Santa Barbara, Ecology, Evolution, and Marine Biology, Santa Barbara, CA, 93106, USA

Keywords:
none specified

Presentation Type: Oral Paper
Number:
Abstract ID:217
Candidate for Awards:None


Copyright © 2000-2019, Botanical Society of America. All rights reserved