Dinitro-o-cresol final screening assessment: chapter 1

Potential to cause ecological harm

A scenario was developed using conservative assumptions to estimate potential environmental concentrations of 4,6-Dinitro-o-cresol (DNOC) that could result from its release from a hypothetical industrial facility. This estimation was done in spite of the absence of specific data indicating releases of DNOC to the environment, in recognition of two factors. First, it is recognized that some industrial facility operators are not fully aware of all potential sources of release of substances to the environment, including rinsate from the cleaning of reactors, transport and storage vessels, either directly at the facility or through third parties such as transporters or container recyclers. Second, there is the possibility that facilities that have not yet been identified are also using the substance. Of note, only facilities that are using greater than 10 000 kg per year of DNOC are required to report to National Pollutant Release Inventory (NPRI). The requirement to respond to the section 71 notice for the year 2000 was also based on a threshold of 10 000 kg.

It was assumed in the conservative scenario that one customer was receiving the total annual import quantity (100-1000 tonnes). The scenario assumed releases of 0.2% of the annual import quantity of DNOC into the St. Clair River; based on professional judgement, this recognizes routine process losses and waste from equipment cleaning for a substance handled in bulk. This accounts for releases to solid waste and wastewater; using this percentage results in an estimated annual release of 200-2000 kg. If it is further assumed that DNOC is in use throughout the year and that there is continuous release (24 hours per day) over the year (350 operating days); daily releases would correspond to approximately 0.57-5.7 kg/day. Sewage treatment plant (STP) removal rates were also considered. The STP model (STP 2001) estimated that 27% of DNOC would be removed and that 73% would enter the environment in the form of final wastewater effluent from an STP.

Two main sources of atmospheric nitrophenols (a category that includes DNOC) have been reported in the literature. These include secondary formation by reactions in the troposphere and emissions from automobiles. Researchers have examined the atmospheric occurrence and formation of DNOC (Nojima et al. 1976; Alber et al. 1989; Richartz et al. 1990). DNOC has been shown to form as a secondary pollutant via the reaction of toluene and 2-methylphenol with nitrogen monoxide and hydroxyl radicals. It is difficult to estimate the quantity that may result from the anthropogenic release of precursor species. Direct emission of DNOC from car exhaust is likely only of minor importance. Under experimental conditions, exhaust from an automobile motor was found to contain DNOC at a rate of < 0.01 ng/m³ (Tremp et al. 1993).

The NPRI (Environment Canada 2003b) reported that amounts of up to 2 tonnes of DNOC and its salts were, prior to 2002, annually “transferred for disposal” by NOVA Chemicals. For all years before 2002, the methods of treatment were biological, such as biooxidation, and incineration or thermal. For the year 2002, disposal was to a landfill.

Ecological effects assessment

Biotic effects

Key studies of the toxicity of DNOC to organisms in different environmental media are presented in tables 6 to 9. Studies primarily on the acute toxicity of DNOC to microorganisms, aquatic invertebrates, insects, terrestrial invertebrates and vertebrates were located in the literature. No acute or chronic marine toxicity data were identified.

EC₅₀ = the concentration of a substance that is estimated to cause some effect to 50% of the test organisms
LC₅₀ = the concentration of a substance that is estimated to be lethal to 50% of the test organisms
NOEC = the no-observed-effect concentration is the highest concentration in a toxicity test not causing a statistically significant effect in comparison to the controls

ED₅₀ = The dose estimated to produce an effect to50% of the population

LC₅₀ = the concentration estimated to be lethal to 50% of the organisms
NOEC = no-observed-effect concentration; LD₅₀ = the dose estimated to be lethal to 50% of the organisms.

LD₅₀ = the dose estimated to be lethal to 50% of the organisms.
LC₅₀ = the concentration estimated to be lethal to 50% of the organisms.
LOEL = lowest-observed-effect level.
CI = confidence interval.

The most sensitive aquatic vertebrates reported in the literature are rainbow trout (Mayer and Ellersieck 1986; Sewell et al. 1995c). The authors reported LC₅₀ values (the concentration estimated to be lethal to 50% of the organisms) of 0.066 and 0.45 mg/L, respectively. The 96-hour LC₅₀ study reported by Sewell et al. (1995c) is an unpublished study; however, it was cited in a peer-reviewed report (IPCS 2000). Atlantic salmon and bluegill are also sensitive, with 96-hour LC₅₀ values of 0.20 mg/L and 0.23 mg/L, respectively (Zitko et al. 1976; Buccafusco et al. 1981).

The effect of DNOC on terrestrial vertebrates (mink and otter) (critical toxicity value [CTV] for wildlife) was calculated using the repeated mammalian (rat) oral dose toxicity data provided for the substance (2.5 mg/kg-bw per day for a 90-day rat dietary exposure study, lowest-observed-effect level [LOEL]) (Den Tonkelaar et al. 1983). The CTV_wildlife is calculated by taking the chronic value (geometric mean of the no-observed-effect level [NOEL] and LOEL) from the rat study and correcting it for body weight of a predictive sentinel species (Sample et al. 1996). In this case, the predictive sentinel species are the piscivorous mammals mink and river otter.

The CTV_wildlife is thus calculated as:

• CTV_wildlife = ChVts · (BW_ts/BW_pss)

where:

• ChV_ts = chronic value for test species (geometric mean of LOEL [2.5 mg/kg-bw per day] and NOEL [0.25 mg/kg-bw per day] = 0.8 mg/kg-bw per day)
  
• BW_ts = mean body weight of test species (0.35 kg)
  
• BW_pss = body weight of predictive sentinel species (0.807 kg for mink; 6.01 kg for otter) (2004 personal communication with Canadian Wildlife Service, Environment Canada, Ontario Region; unreferenced).

Therefore,

• CTV_wildlife = 0.8 × (0.35/0.807) = 0.35 for mink and 0.8 × (0.35/6.01) = 0.047 for otter.

The PNEC_wildlife is calculated from the CTV_wildlife as follows:

• PNEC_wildlife = CTV_wildlife/AF

where:

• PNEC_wildlife = wildlife predicted no-effect concentration (mg/kg-bw per day)
  
• AF = application factor (interspecies variation, laboratory to field extrapolation) (10).

Therefore, PNEC_mink is 0.035 mg/kg-bw per day, and the PNEC_otter is 0.0047 mg/kg-bw per day.

Ecological exposure assessment

Concentrations in the atmosphere and precipitation

No monitoring data for DNOC in the atmosphere or precipitation in Canada were identified. Monitoring data from other countries are summarized in Table 9.

^a Unless otherwise specified. The range of values is indicated in square brackets, if available (for example, [minimum-maximum]).

DNOC has been detected in atmospheric air and precipitation at a number of locations in Europe, and the presence of nitrated phenols in rain is not explained solely by input from pesticide applications (Leuenberger et al. 1988). DNOC has been shown to partition favourably from the gas phase to the aqueous phase, and its presence in rainwater would therefore be expected (Schwarzenbach et al. 2003). DNOC was detected in Denmark, even though the substance had not been used there in the previous 10 years (Danish Environmental Protection Agency 2001). The concentrations found in rain in Denmark are of the same order of magnitude as have been detected in England, and Switzerland.

As no atmospheric or precipitation monitoring data for DNOC in Canada could be located, a series of release scenarios was developed to estimate the amount of DNOC that could be released into receiving waters in Canada as a result of rainfall scavenging of DNOC in the atmosphere. The scenarios incorporated precipitation data for 12 Canadian cities, an estimate of the amount of DNOC in rainwater, and a calculation of runoff from built-up and natural areas into the receiving STPs. It was assumed that the rain event that would result in DNOC being removed from the atmosphere would be a heavy rainfall and that DNOC would be washed out in the early stages of the rain event and not over the length of the rainfall. The concentration of DNOC used in the scenario is based on precipitation values from Europe that were considered realistic possible levels of DNOC in air in Canada. The mean concentration of DNOC in cloudwater from northern Germany (4.2 µg/L) was selected. It was assumed that rainwater would be released as a point source from an STP but that it would not undergo STP treatment, as STP removal efficiency during a storm event is likely to be poor. The highest concentrations of DNOC were estimated in receiving waters from the STPs in London, Ontario (0.0023 mg/L), Guelph, Ontario (0.0023 mg/L), and Granby, Quebec (0.0025 mg/L).

Aquatic concentrations

No recent aquatic monitoring data for DNOC in Canada were identified. Older data on levels of DNOC in Canadian waters as well as in other countries are summarized in Table 10.

^b The range of values is indicated in square brackets, if available (for example, [minimum-maximum]).

^c The range of values is indicated in square brackets, if available (for example [minimum-maximum]).
^d Mean concentration in effluent is presented as an indication of resulting exposure. This value was not included in the section on releases of DNOC, as details on effluent quantities and release rate were not provided.

As no recent Canadian surface water monitoring data were identified, aquatic exposure estimates were modelled. The scenario uses the ChemSim model (Environment Canada 2003c) to predict estimated exposure values. ChemSim model runs were done for three river flow estimates and two loading rates (calculated in the section on releases of DNOC), for a total of six model runs. As indicated in the release scenario, it is assumed that DNOC is in use throughout the year and that there is continuous release (24 hours per day) over the year (350 operating days). Two estimates of low river flow (2.5th and 10th percentiles) were selected to derive predicted environmental concentrations (PECs) under low-flow conditions. The 50th-percentile flow value was also selected to estimate PECs under more typical conditions. The maximum concentration of DNOC at 20 m downstream of the reporting facility with a worst-case scenario release of 5.7 kg/day and a 2.5th-percentile river flow is estimated to be less than 0.006 mg/L. If STP treatment is considered, a PEC of 0.0014 mg/L is estimated.

Concentrations in sediment, sewage sludge and soil

Monitored soil, sediment and sludge concentrations of DNOC are summarized in Table 11. The high flow and velocity of the St. Clair River would rapidly dilute and disperse the substance, and only a minor amount of DNOC is expected to partition to sediments (1%). Based on the results of modelling, at a release rate of 5.7 kg/day, 0.057 kg/day (or 1%) would be available to be adsorbed onto sediments.

^e The range of values is indicated in square brackets, if available (for example, [minimum-maximum]).
^f < W is a qualifier, given to indicate that the sample may contain the analyte but the level would probably not exceed the laboratory method detection limit (MDL). W is approximately one-third to one-fifth of the MDL (OMEE 1994).

DNOC was detected in 13% of Canadian municipal sludges sampled during the period 1980-1985 at concentrations ranging from 1200 to 1500 ng/g dry weight, with a median concentration of 1300 ng/g dry weight (Webber and Lesage 1989). It was not detected (detection limit not stated) in sludge or sludge compost from various locations in Canada sampled in 1993-1994 (Webber and Nichols 1995).

DNOC was not detected (method detection limit = 100 ng/g) in 101 samples of “rural parkland” soil or in 60 samples of “old urban parkland” soil in Ontario (OMEE 1994). Similarly, DNOC was not detected (detection limit = 50 ng/g) in agricultural soil from various locations across Canada (Webber 1994).

Concentrations in biota

DNOC was not detected in fish composite samples (detection limit not stated) from the United States (DeVault 1985).

As indicated in the section on environmental fate and partitioning, DNOC has a relatively low bioaccumulation potential. However, as will be seen in the section on effects characterization, results of repeated oral dose toxicity studies indicate that mammals may be fairly sensitive to DNOC. Therefore, wildlife exposure to DNOC from food and water has been estimated.

A PEC for wildlife was estimated based on a calculation of the total daily intake of the substance by mink and otter. An energetics model based on the general exposure model for wildlife from the U.S. Environmental Protection Agency’s (EPA) Exposure Factors Handbook (US EPA 1993) was used.

where:

• TDI = total daily intake (mg/kg-bw per day)
  
• FMR = normalized free metabolic rate of wildlife receptor of interest (250 kcal/kg-bw per day for mink and river otter)
  
• C_i = concentration of contaminant in the ith prey species (mg/kg-bw) (see below)
  
• P_i = proportion of the ith prey species in the diet (unitless) (default = 35% for mink; 100% for otter)
  
• GE_i = gross energy of the ith prey species (default = 850 kcal/kg-bw prey)
  
• AE_i = assimilation efficiency of the ith prey species by the wildlife receptor (default = 0.91)
  
• Pt = proportion of the time the receptor spends in the contaminated area (= 9% for mink and 0.06% for otter).

The model incorporated the metabolic rate of the wildlife receptors of interest (mink and otter), the proportion of food uptake by the receptors and the amount of time the animals spend in the contaminated area, which is based on the typical habitat range of the wildlife receptors.

The concentration of the substance in a fish (C_i) must be estimated based on the highest PECwater and a BAF. The BAF was estimated using the Modified Gobas Model (Gobas and Arnot 2003). The BAF represents a benthic/pelagic food chain and estimates the accumulation from all sources in a mid-trophic-level fish that would typically be eaten by a mammalian piscivore.

• C_i = PEC_water · BAF

where:

• Ci = concentration in a prey fish (mg/kg-bw)
  
• PEC_water = PEC calculated for surface water (mg/L) (see section on aquatic concentrations)
  
• BAF = bioaccumulation factor for substance (L/kg) (see section on environmental fate and partitioning).
  
• C_i = 0.0014 · 25 = 0.035

The model estimated PECs of 0.0004 mg/kg-bw per day and 0.000 007 mg/kg-bw per day for mink and otter, respectively.