Leaf spectral clusters as potential optical leaf functional types within California ecosystems

Abstract

a b s t r a c tOur ability to measure and map plant function at multiple ecological scales is critical for understanding current and future changes in Earth's ecosystems and the global carbon budget. Conventional plant functional types (cPFTs) based on a few productivity-related traits have been previously used to simplify and represent major differences in global plant functions, but more recent research has directly focused on the use of functional trait information. Still, sampling limitations have constrained efforts to truly understand the variance and covariance of functional traits globally. Reflectance spectra offer a fast, repeatable, simultaneous measurement of a wide variety of leaf functional traits and could be used to optically define leaf functional types. To evaluate this concept, we measured leaf reflectance from a wide range of species in a diverse set of ecosystems across central and northern California, including observations from multiple individuals, sites, and seasons. Using principal components analysis, we analyzed spectral variation in relation to categorical attributes such as species and cPFTs, as well as to a set of functional trait metrics calculated from the spectra. We found the first three principal components (PCs) to be weakly related to categorical attributes and more strongly related to spectrally-derived functional metrics. Each PC was more strongly associated with different portions of the spectrum and contained different functional information. We applied a hybrid clustering algorithm to the PC coordinates of the observations to define potential optical leaf functional types. Twelve spectral clusters were identified, and these did not correspond directly to either single cPFTs or species. However, each cluster had a unique functional metric profile. Clusters represented both inter-and intra-species and cPFT functional differences driven by taxonomy, trait evolution and environmental responses, demonstrating their value as optical leaf functional types and the value of the clustering approach used here for defining optical types from leaf spectra. Our findings support the notion that cPFTs do not adequately capture differences in leaf function. They demonstrate that spectral measurements can be used to improve both the definition of PFTs as well as our knowledge regarding the covariance of functional traits within these classes.

Spectral Properties of Coniferous Forests: A Review of In Situ and Laboratory Measurements

“…Drawing general conclusions about interspecies differences and their magnitude compared to other factors is however difficult, because most studies (Table 1, Appendix A) are local, with only few species and sites measured, and data across campaigns cannot be readily compared, due to methodological differences. Studies that have examined a large number of species, using comparable methodology for each, clearly show that species is important, although not the only factor explaining spectral differences [11,30,33]. For example, for a set of 12 boreal species belonging to five taxonomic genera, species explained up to 71%, 70%, and 78% of variability in the reflectance, transmittance, and albedo spectra, respectively [11].…”

“…For example, for a set of 12 boreal species belonging to five taxonomic genera, species explained up to 71%, 70%, and 78% of variability in the reflectance, transmittance, and albedo spectra, respectively [11]. On the other hand, a study containing 18 plant species (of which 7 were conifers) in California, USA, reported that species explained 20%, 80%, and 47% of variability in the three first principal components derived from the reflectance spectra [33]. In addition, for four temperate and boreal species in North America, it was shown that differences (range) in photochemical reflectance index (PRI) were larger between species than between light conditions [30].…”

Spectral diversity area relationships for assessing biodiversity in a wildland–agriculture matrix

“…For example, if maps of foliar nitrogen concentration and LMA were available, one might find that one of those traits is very random at a certain scale, while the other is very patchy, suggesting something about the environmental vs. biological controls on those two traits at that scale. Since PCs maximize variation and are non‐spatial, we can assume that the first few PCs from hyperspectral data likely correlate with the most variable top‐of‐canopy structural and chemical traits (Roth et al ), but I have not demonstrated that here.…”

A spectral analysis of common boreal ground lichen species

“…Next, we employed a hierarchical clustering analysis to study the grouping of the samples based on their spectral information (e.g. Rees et al, 2004 ; Roth et al, 2016 ). The algorithm starts by assigning each observation to its own cluster and continues with merging the clusters closest in the distance space, finally forming a cluster tree.…”