Reflectance spectroscopy has emerged as a powerful approach for the rapid, non-destructive estimation of plant traits. Plant secondary metabolites (PSMs) are a group of plant traits that are generally under-represented in spectroscopic retrievals of plant functional diversity, even though they play important roles in ecosystem functioning and trophic-level interactions. Moreover, PSMs largely provide the basis for chemical diversity (i.e., chemotypes) within and among plant species. Here we sought to understand the relationships among the diversity of plant phenolic compounds and foliar optical properties using trembling aspen (Populus tremuloides) as a model species.
In 2012 and 2015, foliar reflectance was collected on multiple aspen genotypes at WisAsp (Wisconsin Aspen Plantation), near Madison, WI. In 2012, leaves with reflectance data were assayed for phenolic compounds (i.e., condensed tannins and salicinoids) that compose the secondary metabolites in aspen most influential for ecosystem functioning and an index of phenolic diversity was calculated. We then examined relationships among variability in spectral data and phenolic compounds and built a model predicting the phenolic diversity index as a function of spectral variability. In 2015, we collected replicate foliar reflectance measurements from 79 genotypes at WisAsp to determine the role of secondary metabolites in characterizing aspen genotypes.
Results/Conclusions
We found that variation in spectral profiles was positively related to variation in phenolic profiles. Tannins and the salicinoids tremulacin and salicortin had the strongest relationships with spectral variation. Spectral data could also predict an index of phenolic diversity, with a high model fit and low error of prediction. The accuracy of classification of aspen genotypes using foliar trait data improved when moving from utilizing 1) only primary (e.g., nitrogen, carbon), morphological (e.g., leaf mass per area), and structural data (e.g., fiber, lignin) to 2) including primary traits and an index of phenolic diversity to 3) including primary traits and estimates of specific phenolic traits.
Our findings suggest that not only can reflectance spectroscopy predict specific secondary metabolites that are important for ecosystem functioning, but that variation in foliar optical properties is related with the diversity of those compounds. Incorporating foliar secondary metabolites improved classification of aspen genotypes compared with using only foliar primary, morphological, and structural traits. Understanding the role that the diversity of secondary metabolites plays in the variation of functional traits within a species represents an important step in better integrating functional diversity into ecosystem ecology.