Spectroscopic determination of ecologically relevant plant secondary metabolites. Issue 11 (23rd July 2016)

• Record Type: Journal Article
• Authors: Couture, John J.; Singh, Aditya; Rubert‐Nason, Kennedy F.; Serbin, Shawn P.; Lindroth, Richard L.; Townsend, Philip A.
• Editors: Davey, Matthew
• Is Part Of: Methods in ecology and evolution. Volume 7:Issue 11(2016:Nov.)
• Page Start: 1402
• Page End: 1412
• Publication Date: 2016-07-23
• Subjects: condensed tannins -- paper birch -- phenolics -- phytochemistry -- plant defence -- salicinoids -- salicortin -- spectroscopy -- trembling aspen -- tremulacin; Ecology -- Periodicals; Evolution -- Periodicals; 577
• DOI: 10.1111/2041-210X.12596
• Languages: English
• ISSNs: 2041-210X

Abstract

Summary: Spectroscopy has recently emerged as an effective method to accurately characterize leaf biochemistry in living tissue through the application of chemometric approaches to foliar optical data, but this approach has not been widely used for plant secondary metabolites. Here, we examine the ability of reflectance spectroscopy to quantify specific phenolic compounds in trembling aspen ( Populus tremuloides ) and paper birch ( Betula papyrifera ) that play influential roles in ecosystem functioning related to trophic‐level interactions and nutrient cycling. Spectral measurements on live aspen and birch leaves were collected, after which concentrations of condensed tannins (aspen and birch) and salicinoids (aspen only) were determined using standard analytical approaches in the laboratory. Predictive models were then constructed using jackknifed, partial least squares regression (PLSR). Model performance was evaluated using coefficient of determination ( R 2 ), root‐mean‐square error (RMSE) and the per cent RMSE of the data range (%RMSE). Condensed tannins of aspen and birch were well predicted from both combined ( R 2 = 0·86, RMSE = 2·4, %RMSE = 7%)‐ and individual‐species models (aspen: R 2 = 0·86, RMSE = 2·4, %RMSE = 6%; birch: R 2 = 0·81, RMSE = 1·9, %RMSE = 10%). Aspen total salicinoids were better predicted than individual salicinoids (total: R 2 = 0·76, RMSE = 2·4, %RMSE = 8%; salicortin: R 2 = 0·57, RMSE = 1·9, %RMSE = 11%; tremulacin: R 2 = 0·72, RMSE = 1·1, %RMSE = 11%), and spectra collected from dry leaves produced better models for both aspen tannins ( R 2 = 0·92, RMSE = 1·7, %RMSE = 5%) and salicinoids ( R 2 = 0·84, RMSE = 1·4, %RMSE = 5%) compared with spectra from fresh leaves. The decline in prediction performance from total to individual salicinoids and from dry to fresh measurements was marginal, however, given the increase in detailed salicinoid information acquired and the time saved by avoiding drying and grinding leaf samples. Reflectance spectroscopy can successfully characterize specific secondary metabolites in living plant tissue and provide detailed information on individual compounds within a constituent group. The ability to simultaneously measure multiple plant traits is a powerful attribute of reflectance spectroscopy because of its potential for in situ – in vivo field deployment using portable spectrometers. The suite of traits currently estimable, however, needs to expand to include specific secondary metabolites that play influential roles in ecosystem functioning if we are to advance the integration of chemical, landscape and ecosystem ecology.