Registration Dossier

Environmental fate & pathways

Bioaccumulation: aquatic / sediment

Currently viewing:

Administrative data

Link to relevant study record(s)

Reference
Endpoint:
bioaccumulation in aquatic species, other
Type of information:
calculation (if not (Q)SAR)
Remarks:
Migrated phrase: estimated by calculation
Adequacy of study:
key study
Reliability:
2 (reliable with restrictions)
Rationale for reliability incl. deficiencies:
results derived from a valid (Q)SAR model and falling into its applicability domain, with limited documentation / justification
Remarks:
Values for individual constituents of this natural complex substance (NCS) were calculated using a validated QSAR. All constituents fall within the applicability domain of the QSAR.
Justification for type of information:
See QMRF document "BCFBAF.pdf" attached to this record.
Reason / purpose for cross-reference:
reference to same study
Qualifier:
no guideline followed
Principles of method if other than guideline:
NCSs, consisting of a number of constituents, do not have one single BCF value but a range of BCF values.
The range of BCF is based on calculated values for the individual constituents derived by QSAR from log Kow.
A drawback of every QSAR is that biotransformation is not included. The calculation method for BCF is generally realistic for the actual value, unless biotransformation occurs. When a substance is subject to biotransformation, the actual BCF will be lower than the calculated BCF. Thus, BCF calculations by QSAR are a worst case approach.

The TGD provides the following information about the different QSAR models available for estimating BCF from log Kow: “Among the QSAR models based on the correlation between BCF and Kow, Meylan et al. (1999) compared their proposed fragment-based approach with a linear (Veith & Kosian, 1983) and bilinear (Bintein et al., 1993) model, using a data set of 610 non-ionic compounds. The fragment method provided a considerably better fit to the data set of recommended BCF values than the other two methods, as shown by the higher R2 value, but more importantly, a much lower SD and ME.” This supports the use of the more recently developed model based on Meylan et al (1999) instead of the older model according to Veith et al (1979). The regression method has been revised since 1999 and this QSAR is validated and incorporated in the EPIWIN program as the Log BCF regression-based estimate in the BCFBAF v3.00 programme. Attention is paid that each constituent falls within the applicability domain of the QSAR.
GLP compliance:
no
Key result
Type:
other: log BCF
Value:
1.502 - 2.597 dimensionless
Basis:
whole body w.w.
Calculation basis:
other: QSAR model based on log Kow
Key result
Type:
BCF
Value:
32 - 156 L/kg
Basis:
whole body w.w.
Calculation basis:
other: QSAR model based on log Kow
Remarks on result:
other: for fration of terpenoids, typically ~13% of orange oil
Key result
Type:
BCF
Value:
261 - 395 L/kg
Basis:
whole body w.w.
Calculation basis:
other: QSAR model based on log Kow
Remarks on result:
other: for fraction of hydrocarbon terpenes, typically ~83% of orange oil
Key result
Type:
BCF
Value:
361 L/kg
Basis:
whole body w.w.
Calculation basis:
other: QSAR model based on log Kow
Remarks on result:
other: for limonene, typical 80% of orange oil

Estimated log BCF values for the known constituents of orange oil:

Substance CAS Fraction Estimated log BCF
D-limonene 5989-27-5 0.8 2.557
Citral 5392-40-5 0.01 1.94
beta-Myrcene 123-35-3 0.02 2.418
Linalool 78-70-6 0.04 1.627
Alpha-Terpineol 98-55-5 0.01 1.831
Alpha-pinene 7785-70-8 0.01 2.597
Citronellal 106-23-0 0.01 2.194
Decanal  112-31-2 0.03 2.15
Octanal 124-13-0 0.03 1.502
Conclusions:
The range of log BCF values for the known constituents of orange oil is 1.502 - 2.597. The BCF was in the range of 32 – 156 for oxygenated monoterpenoids, typically 13% of oil and 261 – 395 for hydrocarbon terpenes, typically 83% of oil.
Executive summary:

The log BCF values of known orange oil constituents were estimated using the validated QSAR BCFBAF v 3.00 in the EPIWIN programme. Log Kow estimates were calculated using the corresponding log Kow value estimated by KOWWIN (see chapter 4.7)

The BCF of the constituents in Orange Oil was calculated to be in the range of 32 – 156 for monoterpenoids (typically 13% of oil) and 261 – 395 for hydrocarbon terpenes (typically 83% of oil). As the constituents are readily biodegradable, it is likely that they will also be biotransformed in higher organisms so a high bioaccumulation factor is actually not expected.

Description of key information

The estimated BCF of the constituents in Orange Oil was in the range of 32 – 156 for monoterpenoids, typically 13% of Orange oil and 261 – 395 for hydrocarbon terpenes, typically 83% of Orange oil. With d-limonene as the main constituent the BCF is 361 for >60% (typical 80%) of the NCS.

Key value for chemical safety assessment

BCF (aquatic species):
361 L/kg ww

Additional information

NCSs, consisting of a number of constituents, do not have one single BCF value but a range of BCF values. BCF tests are not adapted to such complex UVCBs such as the NCSs, because of analytical monitoring issues (existing test protocols have been developed for single molecules which are preferably radio-labelled). A constituent approach is followed.

 

The range of BCF can be based on calculated or measured values of the individual constituents. Very few measured BCF values are available for the individual constituents. However, for all constituents, calculated BCF values can be derived by QSAR from log Kow.

 

In the constituent approach, estimated BCF values for the constituents have been used instead of measured. Thus the justification for use of constituent approach and estimated BCF values includes:

  • no BCF data available for Orange Oil
  • BCF difficult to determine for mixtures
  • REACH Annex XI encourages the use of alternative information at all supply levels before a new vertebrate test is conducted
  • the constituents as well as the test substance Orange oil is considered readily biodegradable thus exposure of the aquatic compartment is expected to be low 

A drawback of every QSAR is that biotransformation is not included. The calculation method for BCF is generally realistic for the actual value, unless biotransformation occurs. When a substance is subject to biotransformation, the actual BCF will be lower than the calculated BCF. Thus, BCF calculations by QSAR are a worst case approach. 

 

The TGD provides the following information about the different QSAR models available for estimating BCF from log Kow: “Among the QSAR models based on the correlation between BCF and Kow, Meylan et al. (1999) compared their proposed fragment-based approach with a linear (Veith & Kosian, 1983) and bilinear (Bintein et al., 1993) model, using a data set of 610 non-ionic compounds. The fragment method provided a considerably better fit to the data set of recommended BCF values than the other two methods, as shown by the higher R2 value, but more importantly, a much lower SD and ME.” This supports the use of the more recently developed model based on Meylan et al (1999) instead of the older model according to Veith et al (1979). The regression method has been revised since 1999 and this QSAR is validated and incorporated in the EPIWIN program as the Log BCF regression-based estimate in the BCFBAF v3.00 programme. Attention is paid that each constituent falls within the applicability domain of the QSAR.

Log BCF values of known orange oil constituents were calculated using Log Kow estimates that were calculated using the corresponding log Kow value estimated by KOWWIN (see IUCLID chapter 4.7).

The estimated BCF of the constituents in Orange Oil was in the range of 32 – 156 for monoterpenoids, typically 13% of Orange oil and 261 – 395 for hydrocarbon terpenes, typically 83% of Orange oil. With d-limonene as the main constituent the BCF is 361 for >60% (typical 80%) of the NCS.

As the constituents are readily biodegradable, it is likely that they will also be biotransformed in higher organisms so a high bioaccumulation factor is actually not expected.