142
about the spatial distribution and temporal dynamics of flora and fauna on a relevant
scale (Leutner et al. 2012). In this sense, being able to map and monitor the spatial
distribution of vegetation at species level, and identify changes in population in
space and time, will increase the knowledge about the relationship between ecological processes and ecosystem functioning (Trochet and Schmeller 2013).
Consequently, monitoring species that are introduced or invasive in protected areas
can be applied with greater consistency (Schmeller 2008).
Traditional methods for vegetation mapping, such as exhaustive field surveys,
are both time consuming and costly (Ustin et al. 2004). Remote sensing is now
established as an important tool for researching and monitoring the ecological
processes of terrestrial ecosystems, and has the potential to achieve this in an
efficient and economical way (Nagendra et al. 2010), particularly in sensitive and
inaccessible areas (e.g., mangrove or marshland) (Kamal and Phinn 2011). In the
early years of Earth Observation, some vegetation types and communities were
mapped using broadband multispectral observations, typically from sensors such as
Landsat TM or SPOT (Satellite for Earth Observation). However, with the development of hyperspectral remote sensing, we have the capacity to map at the species
level (Ustin et al. 2004). Among hyperspectral techniques, imaging spectroscopy is
well adapted for airborne platforms and is reinforced with the application of
Remotely Pilot Air Systems (RPAS). However, spaceborne instruments are still in
the early stages of development (Schaepman et al. 2009). In this sense, forthcoming
space missions like EnMAP (Kaufmann et al. 2008) or PRISMA (Stefano et al.
2013) will present a great stimulus to consolidate this approach.
Although airborne imaging spectroscopy with very high spatial and spectral resolution looks promising in the arena of plant species mapping, operational
approaches are lacking because of our limited biophysical understanding of when
remotely sensed signatures indicate the presence of unique species within and
across ecosystems (Somers and Asner 2012). In this sense, two of the main drawbacks
in ecosystems are: high spectral similarity among species with similar ecological
adaptations, and, conversely, high ‘within species’ spectral variability response due
to variations in plant constituents (tissues chemistry and structure) (Asner 1998). To
improve the mapping efficiency of imaging spectroscopy, the analysis techniques
applied to the imagery could be better accomplished if based on ground truth data
to help characterise the spectral response of each plant species (Warner 2010). Field
spectroscopy is the primary method for registering ground truth data to develop
spectral libraries for plants (Manakos et al. 2010). However, to create consistently
unique and detectable spectral signatures among species, this spectral library must
take into account the spatiotemporal variability of the plants, both throughout the
ecosystem and the seasons (Zomer et al. 2009).
Following recommendations of the Rio de Janeiro Convention on Biological
Diversity (CBD) in 1992 and European Habitats Directive (Directive 92/43/EEC),
natural protected sites should be under continuous observation. Consequently, evaluation is undertaken and reporting required every 6 years, to determine the status of
the habitats and species of European importance for nature conservation in a biogeographic region. In this sense, the organizations responsible for the management
M. Jiménez and R. Díaz-Delgado
about the spatial distribution and temporal dynamics of flora and fauna on a relevant
scale (Leutner et al. 2012). In this sense, being able to map and monitor the spatial
distribution of vegetation at species level, and identify changes in population in
space and time, will increase the knowledge about the relationship between ecological processes and ecosystem functioning (Trochet and Schmeller 2013).
Consequently, monitoring species that are introduced or invasive in protected areas
can be applied with greater consistency (Schmeller 2008).
Traditional methods for vegetation mapping, such as exhaustive field surveys,
are both time consuming and costly (Ustin et al. 2004). Remote sensing is now
established as an important tool for researching and monitoring the ecological
processes of terrestrial ecosystems, and has the potential to achieve this in an
efficient and economical way (Nagendra et al. 2010), particularly in sensitive and
inaccessible areas (e.g., mangrove or marshland) (Kamal and Phinn 2011). In the
early years of Earth Observation, some vegetation types and communities were
mapped using broadband multispectral observations, typically from sensors such as
Landsat TM or SPOT (Satellite for Earth Observation). However, with the development of hyperspectral remote sensing, we have the capacity to map at the species
level (Ustin et al. 2004). Among hyperspectral techniques, imaging spectroscopy is
well adapted for airborne platforms and is reinforced with the application of
Remotely Pilot Air Systems (RPAS). However, spaceborne instruments are still in
the early stages of development (Schaepman et al. 2009). In this sense, forthcoming
space missions like EnMAP (Kaufmann et al. 2008) or PRISMA (Stefano et al.
2013) will present a great stimulus to consolidate this approach.
Although airborne imaging spectroscopy with very high spatial and spectral resolution looks promising in the arena of plant species mapping, operational
approaches are lacking because of our limited biophysical understanding of when
remotely sensed signatures indicate the presence of unique species within and
across ecosystems (Somers and Asner 2012). In this sense, two of the main drawbacks
in ecosystems are: high spectral similarity among species with similar ecological
adaptations, and, conversely, high ‘within species’ spectral variability response due
to variations in plant constituents (tissues chemistry and structure) (Asner 1998). To
improve the mapping efficiency of imaging spectroscopy, the analysis techniques
applied to the imagery could be better accomplished if based on ground truth data
to help characterise the spectral response of each plant species (Warner 2010). Field
spectroscopy is the primary method for registering ground truth data to develop
spectral libraries for plants (Manakos et al. 2010). However, to create consistently
unique and detectable spectral signatures among species, this spectral library must
take into account the spatiotemporal variability of the plants, both throughout the
ecosystem and the seasons (Zomer et al. 2009).
Following recommendations of the Rio de Janeiro Convention on Biological
Diversity (CBD) in 1992 and European Habitats Directive (Directive 92/43/EEC),
natural protected sites should be under continuous observation. Consequently, evaluation is undertaken and reporting required every 6 years, to determine the status of
the habitats and species of European importance for nature conservation in a biogeographic region. In this sense, the organizations responsible for the management
M. Jiménez and R. Díaz-Delgado