Effects assessment for the air compartment

Reference can be made to section 5.2.3 and an OECD project in which a testing strategy for terrestrial ecosystems is being developed (Léon and Van Gestel, 1994). Summarising, the assessment factors proposed in Table 16 must be regarded as indicative factors. As more information on the sensitivity of soil organisms becomes available these factors may have to be adjusted.

Table 16 Assessment factors to derive a PNEC

Information available Assessment factor

L(E)C50 short-term toxicity tests

(e.g. plants, earthworms, or micro-organisms)

1000

NOEC for one long-term toxicity test (e.g. plants) 100

NOEC for additional long-term toxicity tests of two trophic levels

50

NOEC for additional long-term toxicity tests for three species of three trophic levels

10

Field data/data of model ecosystems case-by-case

The PNECsoil is calculated on the basis of the lowest effect value measured. If short-term tests with a producer, a consumer and/or a decomposer are available, the test result is divided by a factor of 1000 to calculate the PNECsoil. If only one terrestrial test is available (earthworms or plants), the risk assessment should be performed both on the basis of this terrestrial test and on the basis of the aquatic toxicity data as an indication of the risk to soil organisms. As a precaution, the larger PECsoil /PNECsoil ratio determines which further actions should be taken in the framework of the further testing strategy. The other factors listed in Table 16 are applied, if more tests than the short-term toxicity test have been conducted.

It is clear that the quantitative characterisation of risk by comparison of the PECair to PNECair

is not possible at the moment: only a qualitative assessment for air is feasible.

For the air compartment toxicological data on animal species other than mammals are usually not or only scarcely available. For volatile compounds acute or short-term inhalation tests may be present for new and existing substances. On the basis of these data there may be indications of adverse effects. Short-term LC50 data can be used for a coarse estimation of the risk a chemical poses for animals. However, in most cases, it is unlikely that the atmospheric concentration of a chemical will be high enough to cause short-term toxic effects in the environment, so data on long-term or chronic toxicity should be considered. For example, a chemical may be dangerous for the atmospheric environment at a low concentration, if it is classified as R 48 ("Danger of serious damage to health by prolonged exposure"). Also mutagenic effects and toxic effects on reproduction by a chemical indicate a toxic potential for terrestrial vertebrates.

Fumigation tests on invertebrates are usually not available for new nor for existing substances.

For some existing substances investigations on the toxicity of honey bees (Apis mellifera) which are conducted according to guidelines for the testing of plant protection agents may be available. In these tests, it is sometimes difficult to determine the effective concentration and therefore a PNECair cannot be derived.

Concerning the toxicity for plants, there are almost no data available from tests where a chemical is applied directly via air (gaseous or deposited). Tests with herbaceous species would be desirable but are performed in only a few cases. A guideline for these tests has not been accepted yet.

3.7.2 Abiotic effects

For the evaluation of an atmospheric risk, the following abiotic effects of a chemical on the atmosphere have to be considered:

• global warming;

• ozone depletion in the stratosphere;

• ozone formation in the troposphere;

• acidification.

If for a chemical there are indications that one or several of these effects occur, expert knowledge should be consulted. A first quantitative approach is described in De Leeuw (1993):

Global warming

The impact of a substance on global warming depends on its IR absorption characteristics and its atmospheric lifetime. A potential greenhouse gas shows absorption bands in the so-called atmospheric window (800-1200 nm).

Stratospheric ozone

A substance may have an effect on stratospheric ozone if e.g.

• the atmospheric lifetime is long enough to allow for transport to the stratosphere, and;

• it contains one or more Cl or Br substituents.

In general, ozone depletion potential values approach zero for molecules with atmospheric lifetimes less than one year.

Tropospheric ozone

The generation of tropospheric ozone depends on a number of factors:

• the reactivity of the substance and the degradation pathway;

• the meteorological conditions. The highest ozone concentrations are expected at high temperatures, high levels of solar radiation and low wind speeds;

• the concentration of other air pollutants. The concentration of nitrogen oxides have to exceed several ppb.

Highly reactive compounds (e.g. xylene, olefins or aldehydes) contribute significantly to the ozone peak values. Species with a low reactivity (e.g. CO, methane) are important for ozone formation in the free troposphere and therefore for the long-term ozone concentrations.

However, all studies showed significant variability in the tropospheric ozone building potential values assigned to each organic component. It has to be concluded that at present there is no procedure available to estimate the effect on tropospheric ozone if only the basic characteristics of a substance are known.

Acidification

During the oxidation of substances containing Cl, F, N or S substituents, acidifying components (e.g. HCl, HF, NO2 and HNO3, SO2 and H2SO4) may be formed. After deposition, these oxidation products will lead to acidification of the receiving soil or surface water.

3.8 Assessment of secondary poisoning