473
for individual species from the most to least sensitive. From SSDs, hazard concentrations (HC) affecting a specified percentile of the assemblage are derived, serving
as toxicity benchmark for hazard evaluations. Most commonly a fifth percentile HC
(HC5), or concentration protective of 95% of the assemblage, is computed and
employed in hazard and risk evaluations. However, the generation of SSDs requires
both sufficient number of species and diversity of taxa (e.g., five to greater than ten
aquatic species), which can be hindered by data availability and access to toxicity
and chemistry information. Recent efforts by the response community to facilitate
access to hazard and fate information allows for SSD development for petroleum
and related chemicals. One of these tools is the Chemical Aquatic Fate and Effects
(CAFE) database (NOAA/ER 2015; Bejarano et al. 2016). CAFE currently hosts
aquatic toxicity data for many individual chemical compounds and complex mixtures of chemicals including select crude oils and dispersants.
Despite the availability of information repositories, insufficient toxicity data still
limit the development of SSDs. However, mathematical relationships known as
Interspecies Correlation Estimation (ICE) models have been developed to allow
extrapolation of toxicity between a surrogate species to one or several predicted
species. ICE models have been developed from toxicity data for chemicals with and
without specific modes of toxic action (Willming et al. 2016; Raimondo et al. 2017),
nonpolar narcotics (Bejarano and Barron 2016), and petroleum and dispersant products (Bejarano and Barron 2014). The use of ICE models can augment SSDs comprised of measured data, providing for more robust HC5 estimates (Awkerman et al.
2014).
27.4.2 Deep Ocean Species Sensitivity Distributions
To demonstrate the applicability of SSD and ICE toxicity extrapolation approaches,
datasets for chemicals with at least one deep ocean species were used (Frometa
et al. 2017; Knap et al. 2017; McConville et al. 2018). These data were augmented
with either ICE predicted toxicity values (Bejarano and Barron 2014) based on
hydrocarbon CBB or empirical data queried from CAFE. Data for Corexit 9500
were also extracted from a recent compilation of dispersant-only toxicity studies
(Bejarano 2018). These datasets were used to generate SSDs along with HC5 estimates (Fig. 27.3). SSDs based on CBB for 1-methylnaphthalene and measured total
petroleum hydrocarbon (TPH) concentrations from CEWAF prepared with fresh
crude oil (Alaskan North Slope or Prudhoe Bay plus Corexit 9500) show that deep
ocean species generally occur in the lower end of the sensitivity distribution. In the
case of a larger dataset (Corexit 9500), sensitivities for two deep ocean species were
in the upper portion of the distribution. In all cases, HC5s are protective of the most
sensitive species in the SSD. Although preliminary, these results suggest that SSDs
are useful in visualizing data for deep ocean species of concern. The available data
for deep ocean species were near or above HC5 values indicating the potential protectiveness of HC5 estimates.
27 Ecotoxicology of Deep Ocean Spills
for individual species from the most to least sensitive. From SSDs, hazard concentrations (HC) affecting a specified percentile of the assemblage are derived, serving
as toxicity benchmark for hazard evaluations. Most commonly a fifth percentile HC
(HC5), or concentration protective of 95% of the assemblage, is computed and
employed in hazard and risk evaluations. However, the generation of SSDs requires
both sufficient number of species and diversity of taxa (e.g., five to greater than ten
aquatic species), which can be hindered by data availability and access to toxicity
and chemistry information. Recent efforts by the response community to facilitate
access to hazard and fate information allows for SSD development for petroleum
and related chemicals. One of these tools is the Chemical Aquatic Fate and Effects
(CAFE) database (NOAA/ER 2015; Bejarano et al. 2016). CAFE currently hosts
aquatic toxicity data for many individual chemical compounds and complex mixtures of chemicals including select crude oils and dispersants.
Despite the availability of information repositories, insufficient toxicity data still
limit the development of SSDs. However, mathematical relationships known as
Interspecies Correlation Estimation (ICE) models have been developed to allow
extrapolation of toxicity between a surrogate species to one or several predicted
species. ICE models have been developed from toxicity data for chemicals with and
without specific modes of toxic action (Willming et al. 2016; Raimondo et al. 2017),
nonpolar narcotics (Bejarano and Barron 2016), and petroleum and dispersant products (Bejarano and Barron 2014). The use of ICE models can augment SSDs comprised of measured data, providing for more robust HC5 estimates (Awkerman et al.
2014).
27.4.2 Deep Ocean Species Sensitivity Distributions
To demonstrate the applicability of SSD and ICE toxicity extrapolation approaches,
datasets for chemicals with at least one deep ocean species were used (Frometa
et al. 2017; Knap et al. 2017; McConville et al. 2018). These data were augmented
with either ICE predicted toxicity values (Bejarano and Barron 2014) based on
hydrocarbon CBB or empirical data queried from CAFE. Data for Corexit 9500
were also extracted from a recent compilation of dispersant-only toxicity studies
(Bejarano 2018). These datasets were used to generate SSDs along with HC5 estimates (Fig. 27.3). SSDs based on CBB for 1-methylnaphthalene and measured total
petroleum hydrocarbon (TPH) concentrations from CEWAF prepared with fresh
crude oil (Alaskan North Slope or Prudhoe Bay plus Corexit 9500) show that deep
ocean species generally occur in the lower end of the sensitivity distribution. In the
case of a larger dataset (Corexit 9500), sensitivities for two deep ocean species were
in the upper portion of the distribution. In all cases, HC5s are protective of the most
sensitive species in the SSD. Although preliminary, these results suggest that SSDs
are useful in visualizing data for deep ocean species of concern. The available data
for deep ocean species were near or above HC5 values indicating the potential protectiveness of HC5 estimates.
27 Ecotoxicology of Deep Ocean Spills