187
than large-scale factors such as climate (Bradford et al. 2016). In addition to issues
associated with spatial scale, there are large spans in the scales of biodiversity and
time. Speciose aboveground systems may contain upward of 600 species ha
−1
(Lee
et al. 2002), whereas soils contain many thousands of microbial “species” per gram
of soil, with large numbers of endemics (Schloss and Handelsman 2006). Linking
function to diversity remains a challenge in both systems and particularly in belowground systems where the functional role of the vast majority of species is unknown
(Krause et al. 2014). Likewise, large differences in temporal scales exist between
above- and belowground systems, with leaf responses to sunlight occurring on the
order of seconds (Lambers et al. 1998), while the turnover of soil organic matter can
take years to centuries (Bardgett and van der Putten 2014). Variation in temporal
scales across systems is particularly important given that the importance of biodiversity to ecosystem processes increases with temporal scale (Cardinale et al. 2012;
Reich et al. 2012).
Irrespective of RS, there are shortcomings associated with belowground
measurement. For example, belowground measurements that use enzyme activity
potentials as indicators of microbial function are widespread, but they are known to
have numerous limitations (Nannipieri et al. 2018). Likewise, microbial diversity
estimates based upon amplicon sequences of bacterial 16s rDNA have their own
methodological and interpretive limitations (Schöler et al. 2017). Nonetheless, both
enzyme activities and amplicon sequencing techniques provide useful information
about belowground systems and are used widely enough to be compared across
studies as long as protocols are consistent.
Advantages of using remotely sensed spectral properties of aboveground vegetation to predict belowground processes lie within the data-rich nature of imaging
spectroscopy and the consequent ability to measure many more traits of the canopy
than would otherwise be feasible with traditional benchtop methods. In Madritch
et al. (2014), only four canopy traits were considered using traditional wet chemistry techniques (leaf tannin, N, C, lignin). These canopy foliar traits were expectedly
well correlated with belowground processes. However, plant spectra themselves
were better correlated with belowground processes than were plant leaf traits
(Madritch et al. 2014). This strong relationship between plant spectra and belowground processing existed because the plant spectra provided quantitative information about many plant traits that were not measured via wet chemistry techniques.
Potentially dozens of leaf traits important to belowground processes could be conveyed by plant spectra. The ability of plant spectra to capture many foliar attributes
quickly and accurately is a large reason why plant spectra are useful for predicting
belowground processes. In addition, identifying which regions of plant spectra are
most variable and correlated with belowground process allows researchers to use
spectra to identify plant traits important to soil processes. In short, the potential for
RS products to link above- and belowground systems is promising but faces considerable obstacles.
8 Linking Foliar Traits to Belowground Processes
than large-scale factors such as climate (Bradford et al. 2016). In addition to issues
associated with spatial scale, there are large spans in the scales of biodiversity and
time. Speciose aboveground systems may contain upward of 600 species ha
−1
(Lee
et al. 2002), whereas soils contain many thousands of microbial “species” per gram
of soil, with large numbers of endemics (Schloss and Handelsman 2006). Linking
function to diversity remains a challenge in both systems and particularly in belowground systems where the functional role of the vast majority of species is unknown
(Krause et al. 2014). Likewise, large differences in temporal scales exist between
above- and belowground systems, with leaf responses to sunlight occurring on the
order of seconds (Lambers et al. 1998), while the turnover of soil organic matter can
take years to centuries (Bardgett and van der Putten 2014). Variation in temporal
scales across systems is particularly important given that the importance of biodiversity to ecosystem processes increases with temporal scale (Cardinale et al. 2012;
Reich et al. 2012).
Irrespective of RS, there are shortcomings associated with belowground
measurement. For example, belowground measurements that use enzyme activity
potentials as indicators of microbial function are widespread, but they are known to
have numerous limitations (Nannipieri et al. 2018). Likewise, microbial diversity
estimates based upon amplicon sequences of bacterial 16s rDNA have their own
methodological and interpretive limitations (Schöler et al. 2017). Nonetheless, both
enzyme activities and amplicon sequencing techniques provide useful information
about belowground systems and are used widely enough to be compared across
studies as long as protocols are consistent.
Advantages of using remotely sensed spectral properties of aboveground vegetation to predict belowground processes lie within the data-rich nature of imaging
spectroscopy and the consequent ability to measure many more traits of the canopy
than would otherwise be feasible with traditional benchtop methods. In Madritch
et al. (2014), only four canopy traits were considered using traditional wet chemistry techniques (leaf tannin, N, C, lignin). These canopy foliar traits were expectedly
well correlated with belowground processes. However, plant spectra themselves
were better correlated with belowground processes than were plant leaf traits
(Madritch et al. 2014). This strong relationship between plant spectra and belowground processing existed because the plant spectra provided quantitative information about many plant traits that were not measured via wet chemistry techniques.
Potentially dozens of leaf traits important to belowground processes could be conveyed by plant spectra. The ability of plant spectra to capture many foliar attributes
quickly and accurately is a large reason why plant spectra are useful for predicting
belowground processes. In addition, identifying which regions of plant spectra are
most variable and correlated with belowground process allows researchers to use
spectra to identify plant traits important to soil processes. In short, the potential for
RS products to link above- and belowground systems is promising but faces considerable obstacles.
8 Linking Foliar Traits to Belowground Processes