Registration Dossier

Data platform availability banner - registered substances factsheets

Please be aware that this old REACH registration data factsheet is no longer maintained; it remains frozen as of 19th May 2023.

The new ECHA CHEM database has been released by ECHA, and it now contains all REACH registration data. There are more details on the transition of ECHA's published data to ECHA CHEM here.

Diss Factsheets

Ecotoxicological information

Long-term toxicity to fish

Currently viewing:

Administrative data

Link to relevant study record(s)

Description of key information

Key value for chemical safety assessment

Additional information

In water, the phosphoric - boron trifluoride complex rapidly decomposes to form phosphoric acid and boron trifluoride dihydrate. The latter reacts further to boric acid and fluoroboric acid, which finally hydrolyses to yield hydrofluoric acid/ fluoride ions (BUA, 2005). The assessment of long-term toxicity is therefore based on the products of hydrolysis.

In general, the total phosphorus concentration including phosphate in rivers and lakes reaches levels of several ppm, suggesting that phosphate has no toxic effects (Kim et al. 2012).

The EU Risk Assessment Report for boric acid offers a review of the long-term data available. Longer exposures to boron have been tested, giving chronic or sub-chronic study results. A 34 day study following OECD 210 methods using Brachydanio rerio showed a NOEC value for mortality, growth and condition of 5.6 mg-B/L (Hooftmann et al. 2000). This is the single fish study fully compliant with standard guidelines. However this data is not public. Another study with Brachydanio rerio eggs and embryos showed no toxic effects at exposures less than 10 mmol B/L (Rowe et al., 1998). A NOEC was estimated at 75 mg/L.

Several studies were conducted by Birge and Black (1977) for embryo-larval stages of rainbow trout, goldfish, largemouth bass and channel catfish and are reported as 7 to 9 day embryo-larval stage (ELS) tests. In their 1977 report to their funding agency, Birge and Black reported low level effects to trout (an EC1) at 0.001 mg-B/L, with embryonic mortality and teratogenesis as the endpoints in their 28 day study. This value was an extrapolation from observed data to an unmeasured concentration and was not corrected for control responses. This value was not consistent when the work was repeated. In subsequent work, Black et al. (1993) reported consistent LOEC values at 0.1 mg-B/L for trout in reconstituted water. However, when using waters containing 0.23 to 0.75 mg-B/L, effects were seen only at 1.0 mg added B/L and higher. A longer (87 day) test showed no effects at concentrations up to 18 mg- B/L. Black et al. also collected data on boron concentrations in streams with wild populations of rainbow trout and where trout hatcheries are in operation. They found that wild trout populations survived in streams with 0.01 to 13.1 mg-B/L. They concluded that a concentration of 0.75 to 1.0 mg-B/L appeared to be a “reasonable environmentally acceptable limit” for aquatic systems.

Other reviews pointed out limitations of the 1977 trout values. Guhl (1992) pointed out that the control responses were reported by Birge and Black to be about 12% mortality, but the other data were not adjusted to recognize this baseline. Consequently, reporting a 4-6% response as a significant effect is problematic. Guhl pointed out that the later OECD guidelines permit 30% control mortality, suggesting that smaller responses cannot be relied upon, given the practical limitations of the test. Other studies of field populations showed rainbow trout hatcheries operating inandwith local waters containing up to 0.272 mg-B/L without evident difficulties (Unilever, 1994). Additional information came with demonstration that boron is essential for rainbow trout, zebrafish, and frogs (Eckhert, 1998, Rowe et al., 1998, Fort et al., 1998). The essentiality threshold for rainbow trout appears to be at about 0.1 mg-B/L so tests run with lower concentrations may lead to boron deficiencies.

Because of this and other limitations, the Birge and Black trout studies were rated “Not reliable” and not used in derivation of a PNEC.


Dyer (2001) obtained the original data from the Birge and Black (1977) studies, as well as later data on trout and largemouth bass from Birge and Black (1981) and Birge et al. (1984). Dyer recalculated LC10 values for Oncorhynchus mykiss (trout), Carassius auratus (goldfish), Ictalurus punctatus (channel catfish) and Micropterus salmoide s(largemouth bass). The data for the catfish, goldfish and bass were not as extreme, so have not generated as much controversy and follow-on research. Because the LC10 endpoint is consistent with current practice, the recalculated values were regarded as “reliable with restriction.”