Environmental Engineering Reference
In-Depth Information
are documented (ANZECC and ARMCANZ 2000). In this rating system, weighted
scores are applied to 18 characteristics of the test methodology. The two most heavily
weighted characteristics are exposure duration and endpoint; both must be recorded
for the study to receive a strong rating. Other, lower weighted characteristics, include
control type, organism characteristics, chemical analysis method, exposure type,
test location, chemical grade, test media, hardness/salinity, alkalinity, dissolved
oxygen content, temperature, pH, trend of effect, effect percent, statistical signifi-
cance, and significance level. Based on scores in these categories, data are rated as
C (complete), M (moderate), or I (incomplete). Only data rated C or M are used to
derive guideline values. In addition, the Australia/New Zealand guidelines allow
for use of data that have already been accepted and used in Dutch and Danish water
quality documents. Clear direction on how to handle outlying data is also given in
the Australia/New Zealand guidelines, although the curve-fitting technique used
in this methodology minimizes the need to remove outliers. The Danish methodology
also provides for assessing data quality according to the AQUIRE system (Samsoe-
Petersen and Pedersen 1995).
In addition to the Australia/New Zealand data quality guidelines already discussed,
the ANZECC and ARMCANZ guidelines (2000) provide specific toxicity test validity
criteria. These include that test solutions should cover a geometrically increasing
series; a control and solvent control should be included; control mortalities should
be less than 10% (or some other level, determined by the specific test method);
adverse effects in controls should be less than 20%; water quality parameters
should be measured and be within specified limits; a least significant difference
for hypothesis tests should be calculated and reported; test organisms should be
allowed sufficient time for acclimation to test water; burden of animals in test con-
tainers should be appropriate; measured test concentrations should not vary greatly
from nominal concentrations; animals should be randomly assigned to test vessels,
and test vessels should be randomly placed in test chamber or room; any require-
ments for things such as timing of hatch, or timing and number of young produced
should be met; source and health of test organisms and stock cultures should be
traceable; feeding and no-feeding requirements must be met; reference toxicant
test results for test organisms should be available.
By the Canadian methodology (CCME 1999), each study is evaluated to ensure
that acceptable laboratory practices were used. Studies are classified as primary,
secondary, or unacceptable, with only primary and secondary data being used to
derive guideline values. Primary data must be from toxicity tests conducted according
to currently accepted laboratory practices, but more novel approaches may be
acceptable on a case-by-case basis. Also, for primary data, test concentrations must
be measured at the beginning and end of the exposure period, and static tests are
unacceptable unless test concentrations and environmental conditions were main-
tained throughout the test. Studies should have endpoints from partial or full life
cycle tests and should include determination of effects on embryonic development,
hatching, germination, survival, growth, and reproduction. Appropriate controls
must be included and measurements of abiotic variables (e.g., temperature, and pH)
should be reported. Secondary data may come from tests conducted from a wider
Search WWH ::




Custom Search