109
Another component of Spectranomics is the spectral properties of plant canopies
(Fig. 5.1b). Canopy spectra are derived from the way plant foliage interacts with
solar radiation, and this interaction is strongly determined by foliar chemicals.
Across the full solar spectrum, from the ultraviolet to the visible to the near-infrared
and the shortwave-infrared regions of the electromagnetic spectrum (350–3500 nm),
plants have many common and yet also unique patterns of interaction with solar
energy. Chemometric studies determine how these chemicals relate to reflectance
spectra, and the methods today range from traditional spectroscopic assays and
newer machine learning approaches (Wold et al. 2001; Serbin et al. 2014; Feilhauer
et al. 2015). Spectral properties also provide a tantalizing pathway forward to scale
up from leaves to landscapes (Ustin et al. 2004) to the planetary level (Jetz et al.
2016), but only if we can accurately and repeatedly measure and interpret the spectra of plants over increasingly larger portions of Earth (Fig. 5.1c–e).
The realization of Spectranomics rests in a number of choices made early on to
attempt to reduce unwanted sources of variation combined with extensive sampling.
We focused on humid tropical forests for their high diversity and relative freedom
from extreme phenological changes brought about by seasonal cycles such as those
experienced in temperate regions but may not completely eliminate smaller phenological variation that might arise in reaction to drought or solar variations. We targeted only mature, fully sunlit, top-of-canopy leaves (trees and lianas) to limit
variation attributable to intra-canopy shade and ontogeny and to best relate leaf
properties to airborne and satellite-based spectral measurements. Prior to
Spectranomics, our work and that of many others did not follow a strategically consistent, integrated method for global spectral-functional trait database building
needed to reveal canopy plant functional spectral-chemical patterns at the biospheric scale.
We have collected, cataloged, and stored more than 13,000 canopy tree and liana
specimens, in over 3 million tissue samples, representing about 10,000 species
biased to humid tropical ecosystems (Fig. 5.2a). For perspective, this number
approaches the total number of tree species in the Amazon basin (roughly 11,000;
Hubbell et al. 2008), a value that would put the global tropical tree inventory at
30,000 species if we liberally extrapolate to the entire Neotropics plus the African
and Asian-Oceanic tropics. The Spectranomics database focuses only on species
found in the canopy, meaning they are in full sunlight and are observable from
above. Since roughly 30–60% of tree species in a tropical forest plot makes it to the
canopy (e.g., Bohlman 2015), the current Spectranomics database contains at least
Fig. 5.1 (continued) equations are derived to quantitatively relate canopy functional traits
(chemicals) to spectral data. Example relationships are shown for foliar lignin, nitrogen (N), and
polyphenols. The x-axis indicates spectral wavelengths of 400–2500 nm; the y-axes indicate relative importance of the spectrum to each example chemical constituent shown. (d) An example of
spectra from individual crowns clustered based on their spectral variation. (e) A 3-D view of a
portion of lowland Amazonian forest canopy. Different colors indicate different species detected
based on 15 chemical traits using airborne imaging spectroscopy
5 Lessons Learned from Spectranomics: Wet Tropical Forests
Another component of Spectranomics is the spectral properties of plant canopies
(Fig. 5.1b). Canopy spectra are derived from the way plant foliage interacts with
solar radiation, and this interaction is strongly determined by foliar chemicals.
Across the full solar spectrum, from the ultraviolet to the visible to the near-infrared
and the shortwave-infrared regions of the electromagnetic spectrum (350–3500 nm),
plants have many common and yet also unique patterns of interaction with solar
energy. Chemometric studies determine how these chemicals relate to reflectance
spectra, and the methods today range from traditional spectroscopic assays and
newer machine learning approaches (Wold et al. 2001; Serbin et al. 2014; Feilhauer
et al. 2015). Spectral properties also provide a tantalizing pathway forward to scale
up from leaves to landscapes (Ustin et al. 2004) to the planetary level (Jetz et al.
2016), but only if we can accurately and repeatedly measure and interpret the spectra of plants over increasingly larger portions of Earth (Fig. 5.1c–e).
The realization of Spectranomics rests in a number of choices made early on to
attempt to reduce unwanted sources of variation combined with extensive sampling.
We focused on humid tropical forests for their high diversity and relative freedom
from extreme phenological changes brought about by seasonal cycles such as those
experienced in temperate regions but may not completely eliminate smaller phenological variation that might arise in reaction to drought or solar variations. We targeted only mature, fully sunlit, top-of-canopy leaves (trees and lianas) to limit
variation attributable to intra-canopy shade and ontogeny and to best relate leaf
properties to airborne and satellite-based spectral measurements. Prior to
Spectranomics, our work and that of many others did not follow a strategically consistent, integrated method for global spectral-functional trait database building
needed to reveal canopy plant functional spectral-chemical patterns at the biospheric scale.
We have collected, cataloged, and stored more than 13,000 canopy tree and liana
specimens, in over 3 million tissue samples, representing about 10,000 species
biased to humid tropical ecosystems (Fig. 5.2a). For perspective, this number
approaches the total number of tree species in the Amazon basin (roughly 11,000;
Hubbell et al. 2008), a value that would put the global tropical tree inventory at
30,000 species if we liberally extrapolate to the entire Neotropics plus the African
and Asian-Oceanic tropics. The Spectranomics database focuses only on species
found in the canopy, meaning they are in full sunlight and are observable from
above. Since roughly 30–60% of tree species in a tropical forest plot makes it to the
canopy (e.g., Bohlman 2015), the current Spectranomics database contains at least
Fig. 5.1 (continued) equations are derived to quantitatively relate canopy functional traits
(chemicals) to spectral data. Example relationships are shown for foliar lignin, nitrogen (N), and
polyphenols. The x-axis indicates spectral wavelengths of 400–2500 nm; the y-axes indicate relative importance of the spectrum to each example chemical constituent shown. (d) An example of
spectra from individual crowns clustered based on their spectral variation. (e) A 3-D view of a
portion of lowland Amazonian forest canopy. Different colors indicate different species detected
based on 15 chemical traits using airborne imaging spectroscopy
5 Lessons Learned from Spectranomics: Wet Tropical Forests