Nitrilotriacetic acid

Endpoint summary

Additional information

The EU RAR for trisodiumnitrilotriacetate (EINECS 225 -768 -6), final report, 2005, also contains the relevant ecotox data for nitrilotriacetic acid. This report is attached in section 13. Regarding aquatic toxicity the relevant chapters are quoted below. The tables can be found in the attached full report. Below are the relevant chapters. The key studies have been incorporated in this IUCLID5 dossier as endpoints.

3.2.1 Aquatic compartment (incl. sediment)

A result of the exposure assessment was that in the environment over-stoichiometric amounts of

metal ions are present, thus NTA is always complexed with metal ions. In Section 3.1.3.5, it is

elaborated that always a mixture of different metal complex species occurs in surface waters.

Similar reactions take place in toxicity test media, where metal ions are complexed when

uncomplexed NTA is added. For the interpretation of the test results the complex speciation has

to be considered.

In the effect tests, either H3NTA or the sodium salt was used as test substance. In order to

present comparable results, all effect values are referred as Na3NTA.

3.2.1.1 Toxicity to fish

There is a large database on the toxicity of NTA on fish; an overview of the results considered to

be valid is presented in Table 3.13 and 3.14.

NTA is a readily degradable substance, and in some test media NTA is possibly not stable

throughout the test period. E.g. Macek and Sturm (1973) measured as an average 173 mg/l

Na3NTA in a medium with a nominal concentration of 200 mg/l, and 3.4 mg/l at nominal

5.6 mg/l. Therefore, only studies conducted with analytical monitoring are considered in this

section.

Static and flow-through tests using the bluegill sunfish (Lepomis macrochirus) were conducted

both in soft (60 mg/l CaCO3) and hard water (170 mg/l CaCO3). Monitoring showed essentially

no loss of NTA throughout the test, which was not unexpected as the media were either sterile or

were made from distilled water. The LC50 values in the static test system were 487 mg/l in hard

water and 252 mg/l in soft water, while under flow-through conditions the values were 476 mg/l

and 278 mg/l, respectively. Microscopic examinations of the fish gill tissues revealed slight

damage consisting of the loss of lamellar interdigitation, the beginning of this effect was

observed at 155 mg/l in the hard water static test, 115 mg/l in the soft water static test, 420 mg/l

in the hard water flow-through and 370 mg/l in the soft water flow-through test (Weaver, 1970).

Macek and Sturm (1973) conducted the acute and long-term toxicity of NTA on different fish

species. The test was performed in a flow-through system (detention time 5 hours) with a total

hardness of 35 mg/l CaCO3. 96-hour LC50 values of 98 mg/l for Oncorhynchus mykiss and

127 mg/l for Pimephales promelas were determined. After 28 days of exposure, for Lepomis

macrochirus a LC10 of 96 mg/l and a LC55 of 173 mg/l was determined, while for P. promelas

the LC0 was 96 mg/l and the LC100 173 mg/l. In the long-term test, both species cumulative

mortality due to continuous NTA exposure was conspicuously absent. Examination of gills of

fish exposed to 96 mg/l NTA for 28 days indicated no changes in histology.

Birge et al. (1979) conducted an embryo-larval-test with the fish species channel catfish

(Ictalurus punctatus), goldfish (Carassius auratus), and rainbow trout (Oncorhynchus mykiss,

formerly Salmo gairdneri). Each test was performed in a flow-through system (detention time

2.5 hours) at two water hardness levels (50 and 200 mg/l CaCO3). The NTA concentration was

monitored daily. Exposure was initiated 20 minutes after fertilisation in trout, 1 to 2 hours

post-spawning for goldfish, and 2 to 12 hours after spawning for channel catfish. Average

hatching times were 23, 4.5, and 4 days for trout, catfish, and goldfish. One test parameter was

the egg hatchability, including all embryos (normal or aberrant). Another test parameter was the

survival of normal organisms, determined at hatching and 4 days post-hatching. Normal

organisms were defined as those animals that were free of gross teratic defects. At 4 days

post-hatching, the Na3NTA LC50 values were 90.5, 240.4, and 329.3 mg/l for trout, goldfish, and

catfish stages exposed in soft water, and 114, 243.4, and 384.7 mg/l in hard water. The LC1

values derived by log probit analysis were <16.9 mg/l (trout), 28.5 mg/l (goldfish), and

<131 mg/l (catfish) in soft water, and 20.2, 30.1, and 138.4 mg/l in hard water.

An extended study of the acute and chronic toxicity of NTA on the fathead minnow (Pimephales

promelas) according to the APHA standard procedure was conducted by Arthur et al. (1974).

Water from Lake Superior (hardness about 40 mg/l as CaCO3) was used as the test medium.

Analytical measurements during the test period revealed that NTA was biodegraded (substance

loss up to 24%), therefore all referred concentrations are based on the measurements. The shortterm

toxicity test was conducted in a flow-through test system (retention time 5 hours) with

juvenile fish, the 96-hour LC50 was determined to 114 mg/l Na3NTA. For the chronic

(generation-cycle) test, twenty 3-15-day-old fry were placed in each vessel and exposed to

5 different NTA concentrations (2.1 – 53.9 mg/l Na3NTA). After 30 days exposure, larval

growth was not affected even by the highest tested concentration. After an exposure period of

224 days, there were no observable differences in survival, spawning activity, and egg

hatchability at the highest tested concentration of 53.9 mg/l Na3NTA. At this concentration,

NTA is mainly complexed with Ca and Mg. The individual exposure time of single development

stages is not definitively specified.

Tests on acute toxicity to fish resulted in 96-hour LC50 values in the range of 98 – 487 mg/l. In

all of these tests effects were observed when the Na3NTA concentration exceeded the

stoichiometric metal levels (mainly Ca and Mg) in the medium. A generally accepted hypothesis

is that the toxicological profile of complexing agents is based on disturbances of metal

metabolism. It is expected that effects are caused by the uncomplexed agent. This is supported

by the increased effect values in hard water. Even in the 28-day test with adult fish (Macek and

Sturm, 1973) the LC0 and LC10 values of 96 mg/l are approximately equal to the stoichiometric

metal levels.

Lower effect values (LC1 in the range of <16.9 – <131 mg/l) were determined by the

embryo-larval tests conducted by Birge et al. (1979). The effect values are usually very low

compared to effect values found by other authors. No explanation for these discrepancies could

be found. A careful examination of the entire information provided by Birge et al. gave no

plausible reason for the inconsistency of the data. However, as it was not possible to reproduce

the effect values, it was decided by the EU member states not to use these data for a derivation of

a PNECaqua if other valid fish early life stage tests are available. Therefore, the effect values

found by Birge et al. are not employed in the further effects assessment.

In a generation-cycle test over 224 days on Pimephales promelas (Arthur et al., 1974), there

were no observable differences in survival, spawning activity, and egg hatchability at the highest

tested concentration of 54 mg/l Na3NTA (the active test substance was Ca- or Mg-NTA). Based

in this study, the NOEC for fish is determined to 54 mg/l.

3.2.1.2 Toxicity to Invertebrates

Because of the ready degradability of NTA, in static tests the test solutions are probably not

stable over the total test period. Based on the available monitoring data, it is expected that the

stability is guaranteed up to 48 hours. Static or semi-static tests over a longer period are possibly

not valid and thus not referred here. There is a large database on the toxicity of Na3NTA on

invertebrates; an overview of the results considered to be valid is presented in Table 3.15 and

3.16.

a) Crustacea

An immobilisation test on Daphnia magna in a medium with a water hardness of 286 mg/l

CaCO3 was conducted by Bringmann and Kühn (1977a). It is not clear whether H3NTA or

Na3NTA was used as test substance. The test solution was neutralised (pH 7.6 – 7.7). After

24 hours exposure, an EC0 of 800 mg/l, an EC50 of 950 mg/l, and an EC100 of 1,350 mg/l were

obtained. In a further test (Bringmann and Kühn, 1982) in a similar medium without

neutralisation, the effect concentrations were EC0 75 mg/l, EC50 79 mg/l, and EC100 83 mg/l.

After neutralisation (pH 8.0), the EC50 was above 1,000 mg/l. The test substance was not

monitored, thus all concentrations are nominal. The results indicate that the effects were largely

due to the change of the pH value.

A static short-term toxicity test with the crustacea Daphnia magna was carried out in a medium

with a hardness of 220 mg/l CaCO3 (Canton and Sloof, 1982). The EC50 value was in the range

of 560 – 1,000 mg/l for Daphnia (endpoint: immobilisation, mortality).

Flannagan (1971) tested the toxicity of Na3NTA on 17 species of macro-invertebrates using

4 different natural waters with different hardness. Monitoring of the test substance revealed no

significant decrease over a period of 73 hours. Hyallela azteci was tested in a flow-through

system in unbuffered water (pH 9.3, 21 mg/l CaCO3), the 72-hour LC50 was above 250 mg/l.

Experiments with Gammarus lacustris in a static system showed a LC50 of about 600 mg/l in

unbuffered hard water. Tests with Pontoporeia affinis in a flow-through system, the LC50 was

above 1,000 mg/l in buffered soft water (21 mg/l CaCO3)

An extended study of the acute and chronic toxicity of NTA on the amphipod Gammarus

pseudolimnaeus according to the APHA standard procedure was conducted by Arthur et al.

(1974). Water from Lake Superior (hardness about 40 mg/l as CaCO3) was used as the test

medium. Analytical measurements during the test period revealed that NTA was biodegraded

(substance loss up to about 50%); therefore all referred concentrations are based on the

measurements. Both tests were conducted in a flow-through test system with a retention time of

5-6 hours. The 96-hour LC50 was determined to 98 mg/l Na3NTA. For the chronic test,

25 individuals of 18-day-old newly hatched young were placed in each vessel and exposed to

5 different NTA concentrations (1.2 – 51.9 mg/l Na3NTA). Over 21 weeks exposure, the

recorded parameters were survival, final gravid females, number of produced young, total births,

number of young per female, and births per female. No young or births were produced in the

18.7 and 51.9 mg/l test chambers. The reproduction index (determined by adding total births plus

final gravid females divided by final number of surviving females) showed a significant decrease

in female fecundity in concentrations ≥ 18.7 mg/l. The lowest tested concentration without

significant effects was 9.3 mg/l Na3NTA. At this concentration, NTA is mainly complexed with

Ca and Mg.

b) Molluscs

Weaver (1970) conducted short-term tests with the snail Physa heterostropha in both soft and

hard water. 10 adult snails with an average diameter of 1.0 cm were exposed for 96 hours.

Monitoring showed essentially no loss of NTA throughout the test, which was not unexpected as

the media were either sterile or were made from distilled water. The LC50 values were 522 mg/l

in hard water (170 mg/l CaCO3) and 373 mg/l in soft water (60 mg/l CaCO3).

Flannagan (1971) tested the toxicity of Na3NTA on 17 species of macro-invertebrates using

4 different natural waters with different hardness. Monitoring of the test substance revealed no

significant decrease over a period of 73 hours. Helisoma trivolis was tested in both buffered and

unbuffered media, in both experiments the LC50 was above 250 mg/l. At higher Na3NTA

concentrations, the toxicity was higher in the unbuffered medium; the lethal effects are probably

due to the increase of pH. Experiments with Physa sp. resulted in LC50 values of about 400 mg/l

in a water hardness of 21 mg/l CaCO3 and about 700 mg/l in a water hardness of 745 mg/l

CaCO3.

A test on the influence of NTA on mortality, growth and fecundity through 4 generations of the

freshwater snail Helisoma trivolis was conducted by Flannagan (1974). 10 juvenile snails were

exposed to 5 Na3NTA concentrations in a flow-through system, their weight was measured daily

during the 120-day exposure period. 10 snails of the offspring of each concentration group were

selected to form the next generation. Monitoring of NTA revealed that the test substance

concentration was stable. No significant growth differences were found between snails exposed

to 6.25 and 12.5 mg/l Na3NTA and their controls. At 25 mg/l only the F1 snails were

significantly smaller, while with 50 mg/l the F1 and F2 snails but not the F3 generation was

smaller than the control groups. At 100 mg/l all four generations were reduced in weight. Lethal

effects were not observed up to 100 mg/l. From this test, a NOEC of 12.5 mg/l can be

determined.

c) Insects

A static short-term toxicity test with 3-4 weeks old larvae of the insect Aedes aegypti was carried

out in a medium with a hardness of 220 mg/l CaCO3 (Canton and Sloof, 1982). The LC50 was in

the range of 5,600 – 10,000 mg/l.

All tests on acute toxicity to invertebrates showed effects only when the Na3NTA concentration

exceeded the stoichiometric metal levels of the medium. It is expected that effects are caused by

the uncomplexed agent. This is supported by the increased effect values in hard water.

In long-term tests, the most sensitive organism was the amphipod Gammarus pseudo limnaeus.

In a generation-cycle test over 21 weeks exposure, the lowest tested concentration without

significant effects was 9.3 mg/l Na3NTA. Based in this study, the NOEC for invertebrates is

determined to 9.3 mg/l. At this concentration, NTA is mainly complexed with Ca and Mg.

3.2.1.3 Toxicity to algae

The influence of medium composition on the growth inhibition of 3 algal species (Selenastrum

capricornutum, Scenedesmus subspicatus, Chlorella vulgaris) was examined by Millington et al.

(1988). Bolds Basal medium (BBM) is a very rich medium containing much higher

concentrations of nutrients compared to OECD and EPA media. The method used followed the

OECD test guideline. NTA (unclear whether acid or sodium salt) was tested at 5, 10, 50, 80, and

100 mg/l. The 5-day NOECs (related to cell concentration) are 5 mg/l for all 3 species in both

OECD and EPA medium, while 50 mg/l (S. capricornutum) and 80 mg/l (S. subspicatus, C.

vulgaris) for BBM was obtained. The test results indicate that the apparent effects are mainly

caused by nutrient deficiency.

Both static and continuous flow tests on growth inhibition of the diatom Navicula seminulum

using hard and soft nutrient solution was conducted by Weaver (1970). Test cultures were

prepared by placing diatom stock solution onto millipore filters and introducing the filters into

flasks containing nutrient solutions. At the conclusion of each test the cultures were dried and

weighed. Monitoring throughout the tests showed essentially no loss of NTA. In the static test,

the 96-hour ECb50 were 477 mg/l for hard water and 185 mg/l for soft water. Similar results were

obtained in the flow-through system, the 96-hour ECb50 were 477 mg/l for hard water and

133 mg/l for soft water. In both media, the concentrations of nutrient metals (e.g. 2 mg/l ZnSO4

or 1 mg/l CoCl2) were relatively high thus preventing nutrient deficiency.

A static growth inhibition test on Chlorella vulgaris and Microcystis aeruginosa was conducted

by Canton and Slooff (1982). The 96-hour ECb50 for C. vulgaris is in the range of

560-1,000 mg/l and for M. aeruginosa in the range of 180 – 320 mg/l Na3NTA. The

concentrations of nutrient metals (e.g. 110 μg/l ZnCl2 or 80 μg/l CuSO4) in the test medium were

relatively high thus preventing nutrient deficiency.

The effect of NTA on the marine algae Cyclotella nana using synthetic seawater was studied by

Erickson et al. (1970). Growth rates (determined as cell density) were determined with 0.25, 0.5,

1.0, 2.5, and 5.0 mg/l Na3NTA after 72 hours. The nutrient metal concentrations are comparable

with the OECD standard medium. Addition of 0.25 to 0.5 mg/l Na3NTA resulted in greater

growth at 8.5°C, but not at 2.0°C. The addition of 1.0 to 5.0 mg/l resulted in a progressive

inhibitory effect with time (EC50 about 2.5 mg/l) at both temperatures. The inhibitory effect is

attributed to the reduced bioavailibility of trace metals.

The apparent toxicity of complexing agents to algae in standard tests is related to essential trace

metal bioavailibility. Trace metal levels tend to be more important in algal growth tests than in

other short-term tests (e.g. on fish or daphnia); the main reason is the rapid increase of biomass

during the test. In standard tests using uncomplexed agents, the concentrations of free essential

metal ions decrease drastically, leading to nutrient deficiency and relatively low effect

concentrations. Addition of higher amounts of nutrient metals result in detoxification of the

agent. This effect is also known from other complexing agents, e.g. EDTA (Dufková, 1984) and

[S,S]-Ethylenediamine disuccinate, [S,S]-EDDS (Schowanek et al., 1996), with both substances

the apparent toxicity disappeared when stoichiometric amounts of the nutrient metals were

added.

3.2.1.6 Toxicity to Microorganisms

A series of monospecies tests to microorganisms is available; an overview is presented in Table 3.18.

a) Bacteria

A static growth inhibition test on Pseudomonas fluorescens was conducted by Canton and Slooff

(1982). The tested bacteria were in the log phase at the start of exposure. The medium contained

5,000 mg/l glucose as substrate. The 96-hour EC50 values are in the range of 3,200 – 5,600 mg/l.

Bringmann and Kühn (1977b) tested the inhibition of cell multiplication with Pseudomonas

putida. No effects were observed in concentrations up to 10,000 mg/l H3NTA (= 13,500 mg/l

Na3NTA) after 16 hours of exposure.

b) Protozoa

A test on cell multiplication inhibition with different protozoa was performed using identical

experimental conditions. Stock and preliminary cultures of the test organisms were fed with

living bacteria, whereas the test cultures were fed with inactivated bacteria. H3NTA was used as

test substance. After 72 hours exposure, the toxic threshold concentrations (EC5) were

> 400 mg/l H3NTA (> 540 mg/l Na3NTA) for Chilomonas paramaecium (Bringmann et al.,

1980), > 800 mg/l H3NTA (> 1,100 mg/l Na3NTA) for Entosiphon sulcatum (Bringmann, 1978),

and > 800 mg/l H3NTA (> 1,100 mg/l Na3NTA) for Uronema parduzci (Bringmann and Kühn,

1980).