Predicting leaf traits across functional groups using reflectance spectroscopy

Abstract

Summary Plant ecologists use functional traits to describe how plants respond to and influence their environment. Reflectance spectroscopy can provide rapid, non‐destructive estimates of leaf traits, but it remains unclear whether general trait‐spectra models can yield accurate estimates across functional groups and ecosystems. We measured leaf spectra and 22 structural and chemical traits for nearly 2000 samples from 103 species. These samples span a large share of known trait variation and represent several functional groups and ecosystems, mainly in eastern Canada. We used partial least‐squares regression (PLSR) to build empirical models for estimating traits from spectra. Within the dataset, our PLSR models predicted traits such as leaf mass per area (LMA) and leaf dry matter content (LDMC) with high accuracy (R2 > 0.85; %RMSE < 10). Models for most chemical traits, including pigments, carbon fractions, and major nutrients, showed intermediate accuracy (R2 = 0.55–0.85; %RMSE = 12.7–19.1). Micronutrients such as Cu and Fe showed the poorest accuracy. In validation on external datasets, models for traits such as LMA and LDMC performed relatively well, while carbon fractions showed steep declines in accuracy. We provide models that produce fast, reliable estimates of several functional traits from leaf spectra. Our results reinforce the potential uses of spectroscopy in monitoring plant function around the world.

Corrigendum to Predicting leaf traits across functional groups using reflectance spectroscopy

“…The authors used the internal validation procedure described in Kothari et al . (2023b) to train prediction models for fresh‐leaf EWT based on reflectance, transmittance, and absorptance spectra. The retrained models showed similar but (as predicted) slightly better performance compared with the original models for rehydrated EWT.…”

“…

Since its publication, the authors of Kothari et al (2023b) have identified an error in the Supporting Information published alongside their article. In the Supporting Information Methods S1 section, it is stated that both the fresh and the rehydrated mass of all leaf samples were measured.

…”

“…Given these considerations, the authors trained new models to predict EWT based on the fresh mass rather than the rehydrated mass.Fresh-leaf and rehydrated EWT values were very strongly correlated (R 2 = 0.971) because the authors' protocol focused on sampling healthy mature green leaves without visible signs of stress, including water stress (Lalibert e, 2018). Nevertheless, fresh EWT was generally lower because most leaves were not fully hydrated under field conditions (median ratio: 87.7%, 2.5-97.5 th percentile: 68.3-97.6%).The authors used the internal validation procedure described in Kothari et al (2023b) to train prediction models for freshleaf EWT based on reflectance, transmittance, and absorptance spectra. The retrained models showed similar but (as predicted) slightly better performance compared with the original models for rehydrated EWT.…”

Innovations in plant science from integrative remote sensing research: an introduction to a Virtual Issue

“…Hyperspectral instruments are capable of detecting changes in leaf reflectance that are associated with variation in a suite of functional traits, including leaf physical structure; concentrations of pigments, nutrients, nonstructural carbohydrates, and leaf water status (Wang et al ., 2020). The performance and generality of spectroscopy‐based methods varies among function traits – the studies featured here, published in New Phytologist , demonstrated promise in predicting leaf carboxylation capacity, leaf mass per area, and leaf age with spectroscopy‐based models that generalize across sites, species, and canopy positions in tropical forests (Chavana‐Bryant et al ., 2017; Wu et al ., 2017, 2019); across multiple ecological domains in the eastern United States and Canada (Wang et al ., 2020; Kothari et al ., 2023); and across biomes, including Arctic and tropical sites (Serbin et al ., 2019). Furthermore, Yan et al .…”

“…newphytologist.com/virtualissues.traits, including leaf physical structure; concentrations of pigments, nutrients, nonstructural carbohydrates, and leaf water status(Wang et al, 2020). The performance and generality of spectroscopybased methods varies among function traitsthe studies featured here, published in New Phytologist, demonstrated promise in predicting leaf carboxylation capacity, leaf mass per area, and leaf age with spectroscopy-based models that generalize across sites, species, and canopy positions in tropical forests(Chavana-Bryant et al, 2017;Wu et al, 2017Wu et al, , 2019; across multiple ecological domains in the eastern United States and Canada(Wang et al, 2020;Kothari et al, 2023); and across biomes, including Arctic and tropical sites(Serbin et al, 2019). Furthermore,Yan et al (2021) demonstrated that spectroscopy-based methods are able to predict leaf carboxylation capacity more accurately, and with more generality across sites, than correlations with a suite of other functional traits.…”