Microtox ®

Acute Toxicity Manual

MODERN WATER
A Deepverge Company
 Microtox® - Acute Toxicity Manual

Contents

Section 1 - General Information Page No.

Fundamentals page 3
Supplies and accessories page 4
Sample collection and preparation page 8
Test versatility page 9
Section 2 - Applications and methods

Section 3 - Reference standard information

Phenol standard
Zinc sulfate standard
Section 4 - Data reduction formulas
Basic test protocols
Solid phase protocols
Comparison test
Section 5 - Special handling

Section 6 - FAQ’s

MATERIALS REQUIRED

• Microtox® Acute Toxicity Test Reagent

• Reconstitution solution

• Osmotic Adjusting Solution (OAS)

• Diluent

• Cuvettes

• 10 - 100 μL pipettor and tips

• 0.25 - 2.5 μL pipettor and tips

• Repeat pipettor and syringes

2
Section 1 MODERN WATER
A Deepverge Company

Basic Test Fundamentals

Principle of Operation Time and Temperature
The test exposes luminescent test organisms Different chemicals affect living organisms
in Microtox Acute Toxicity Test Reagent to test at different rates, reflecting differences in
samples, and measures the increase or mechanism of action.
decrease in light output by the test organisms.
For some classes of chemicals, the effect on
Reagent contains living luminescent light output is complete in 5 minutes. For
bacteria that have been grown under other classes of chemicals, the light output is
optimal conditions, harvested, and then still decreasing rapidly at 5 minutes. In these
lyophilized (freeze-dried). The lyophilized cases, 15-minute data may be more reliable.
bacteria are rehydrated with Reconstitution We recommend that 5-minute and 15-minute
Solution to provide a ready-to-use data be taken routinely when dealing with
suspension of organisms. unknown samples. Some laboratories elect
to collect data routinely at five, fifteen and
The test system measures the light output of
thirty minutes. The software will automatically
the luminescent bacteria after they have
process data taken at any one, two, or three
been challenged by a sample and compares
times you select.
it to the light output of a control (reagent
blank) that contains no sample. A difference The temperature during exposure to various
in light output (between the sample and the materials will affect the response of the
control) is attributed to the effect of the living organisms. For the Microtox Acute
sample on the organisms. Toxicity Test procedures, the Incubator Wells
and READ Well temperature are 15°C and the
Precision REAGENT Well temperature is 5.5°C.

Each test cuvette contains roughly a million Control (Blank)

individual test organisms that are challenged
by the test sample. Variations among Reagent Control is required for each test and
individual organisms become statistically is run concurrently with the test. Light levels
insignificant. The system measures a single normally change with time, for reasons other
parameter, the simultaneous light output of than the bioreactivity of the test sample.
all of the organisms.
Use of the Reagent Control allows the
Each batch of Reagent, containing lyophilized operator to compensate for some of these
test organisms, is prepared under conditions variables when reducing the data. In all
within very rigidly controlled limits. variations of the Basic Test procedure, the
responses of all the test cuvettes are
normalized to that of the Control test
response, which compensates for errors up
to 20% in the 10 μL pipetting of reagent.

3
 Microtox® - Acute Toxicity Manual Section 1

Supplies and Accessories

Testing requires several special items in addition to those commonly found in testing laboratories.

Reagent Diluent
The Microtox Acute Toxicity Test Reagent is The Diluent is a specially prepared nontoxic
specially formulated for bioreactivity testing 2% Sodium chloride (NaCl) solution, used for
with sensitivity to a broad range of toxicants. diluting the sample and the reagent.
The reagent is a freeze-dried preparation The marine bacterium in the reagent requires
of a specially selected strain of the marine osmotic protection that is provided by the
bacterium Aliivibrio fischeri (formerly known 2% NaCl.
as Vibriio fischeri, NRRL number B-11177). A vial
of reagent contains roughly one hundred Shelf life is three years when stored at
million test organisms. room temperature.

Shelf life is two years when stored at minus Osmotic Adjustment Solution
(-) 20ºC to (-) 25ºC. Do Not store the Reagent
below minus (-) 25ºC. Osmotic Adjustment Solution (OAS) is a
specially prepared nontoxic 22% Sodium
Storage at normal refrigerator temperature chloride (NaCl) solution, used to adjust the
greatly reduces shelf-life. osmotic pressure of the sample to
Reagent should not be stored in a approximately 2% NaCl.
self-defrosting freezer, which defrosts
To adjust a sample osmotically, add one part
by warming up periodically.
of OAS to ten parts of sample.
Use the reagent within three hours after
reconstitution. The sensitivity of the reagent (X mL sample x 0.1 = mL OAS)
is essentially unchanged for 3 hours after Example: 2.5 mL sample x 0.1 = 0.25 mL OAS
reconstitution. Changes in sensitivity may
become significant after that time for The sample can also be osmotically adjusted
some samples. by the addition of solid AR grade NaCl to a
final concentration of 2.0%.
Reconstitution Solution
Shelf life is three years when stored at
Reconstitution Solution is a specially prepared room temperature.
nontoxic Ultra Pure Water that contains trace
amounts of sodium chloride (NaCI). DO NOT PREPARE SOLUTIONS
Shelf life is three years when stored at Do not make Diluent, Osmotic Adjustment
room temperature. Solution or Reference Water yourself or
use substitutes. The production of
uncontaminated solutions is difficult,
and some laboratories have created
problems using their own Diluent and
Osmotic Adjustment Solution.

4
Section 1 MODERN WATER
A Deepverge Company

Cuvettes
Cuvettes are used to contain samples,
controls, and Reagent during testing. They are
nontoxic and disposable.
Do NOT reuse cuvettes. Used cuvettes cannot
reliably be cleaned for reuse. Traces of
detergent or sample contaminants interfere
with later tests. The risk of interference from
contamination is unacceptably high.

Pipettors and Pipettor Tips

Test protocols require repeated precise
transfer of small amounts of liquid, as
little as 10 μL. High precision adjustable
micro-pipettors are necessary. It is highly
recommended to have and use the
following pipettors:

10 - 100 μL adjustable volume pipettor

0.25 - 2.5 mL adjustable volume pipettor
repeat pipettor

5
 Microtox® - Acute Toxicity Manual Section 1

Sample Collection and Preparation

Sample Collection Most of the time when either a Municipal
Water Treatment or Waste Water Treatment
Use new clean, borosilicate, screw cap Plant wants to check for acute toxicity, they
containers (30 to 50 mL) with Teflon® or do not want to know the effect of chlorination.
equivalent lined caps. Fill the container
Collect the samples before chlorination,
completely to the top with sample, leaving
unless you are testing for the effect of
NO airspace. Completely filling the container
chlorination. Do not collect a chlorinated
helps keep volatile material in solution.
sample for testing, unless you want to know
Sample Storage the effect of chlorination or can not collect
a sample that is not chlorinated. If knowing
Test the sample as soon as possible after the effect of chlorination is desired, it is
collection. If testing is delayed, store samples recommended to test the sample just before
at normal refrigerator temperature. chlorination and just after chlorination.
When the effect of chlorination can not be
The toxicity of the sample can change with avoided, dechlorinate the sample with
time, so testing the sample within 1-2 hours Sodium Thiosulfate.
after collection is best, but this is not always
possible. Try to test the sample within 24-48 Sodium Thiosulfate Dechlorination
hours after collection.
To dechlorinate a sample, add 1 part
Colored/ Turbid Sample of Sodium Thiosulfate Stock Solution to
100 parts of Sample.
The Microtox LX performs automatic color
correction. This feature can be turned off if Example: 100 uL Sodium Thiosulfate Stock
color correction is not desired. Solution added to 10 mL Sample.
Final Sodium Thiosulfate concentration in
When color correction is turned on and the the sample is 100 mg/L.
sample is too colored or too turbid, the
software will generate a message to the Sodium Thiosulfate Stock Solution Preparation
user and an initial dilution may be required.
1. Weight out 1.0g Sodium Thiosulfate
The turbidity is objectionable when the toxicity (Na2S2O3).
from the turbid material is not wanted. If the
2. Add 100.0 mL Diluent.
toxicity is desired from the material that
causes turbidity, do not centrifuge the sample The Sodium Thiosulfate will last for 1-2 months
and ensure that "measure no turbidity" is when stored in the refrigerator.
chosen on the analyzer.

Chlorine Content
The Reagent (organism) is sensitive to
chlorine as are all microorganisms.
Municipal Water Treatment (drinking water)
or Waste Water Treatment Plants have a
common problem with the water they are
producing, bacterial contamination.
Chlorination of the water is generally used
to solve this problem. Chlorine is toxic: that is
why it is being used for killing microorganisms.

6
Section 1 MODERN WATER
A Deepverge Company

Sample pH
1. Measure the pH of the sample and record
the pH as part of the information in the
sample description.

The bacterial reagent is sensitive to pH.

There is a minimal pH effect between 6.0
and 8.0. When the pH is higher than 8.0 or
lower than 6.0, and the sample has buffering
capacity, the effect can be dramatic.
Example: distilled water adjusted to pH 5.0
with HCl has NO pH effect because there are
not enough hydrogen ions to affect the pH
of the Reagent. Another sample at the same
pH (5.0) may show considerable toxicity if it
does have buffering capacity.
2. If the pH of the sample is below 6.0 or
above 8.0 and pH adjustment is required,
adjust the pH as shown below.
Adjustment is required when the toxic
effect of pH is not wanted.

A. If the pH is below 6.0, adjust the pH to

6.0 with NaOH. If over-titration occurs,
discard sample and start again.
B. If the pH is above 8.0, adjust the pH to
8.0 with HCl. If over-titration occurs,
discard sample and start again.
NOTE: When the sample is below pH 6.0,
titrating the sample above 6.0 may
precipitate the sample, thus changing the
effective toxicity.
When the sample is above pH 8.0, titrating
the sample below 8.0 may precipitate the
sample, thus changing the effective toxicity.
When the sample has been over-titrated,
back titrating may not resolubilize the
sample, if precipitation occurred. Use a
strong concentration of acid or base (5 N)
for coarse pH adjustment of a sample with a
high/strong buffering capacity to minimize
sample dilution.
Use a weak concentration of acid or base
(0.5 N) when the sample has a low/weak
buffering capacity.
7
 Microtox® - Acute Toxicity Manual Section 1

Sample Collection and Preparation

The table below shows the combinations of numbers of tests, controls and dilutions that can
be performed at one time. The Basic Test protocol with slight modifications can take readings
on one to three controls, plus three to fourteen serial dilutions of the same sample (a combined
maximum of fifteen cuvettes).

TEST VERSATILITY

No. Tests No. Controls No. Dilutions

1 1 4-14

1 2 4-13

1 3 4-12

2 1 4

3 1 4

ID 1 4-9

ID 2 4-8

ID 3 4-7

2D 1 4

D = Test in duplicate

8
Section 2 MODERN WATER
A Deepverge Company

Applications & Methods

Wastewater treatment Mining
*Influent (inc trucked waste): Basic, 2% Basic, Basic, 2% Basic, 2% Screening
2% Screening, In process: Basic
* Applications available
*Effluent: 81.9% Basic, Comparison, WET,
ISO, DIN Microtox Methods
*TRE: Basic, Comparison
There are 19 methods on the Microtox LX
Drinking water software and 17 standard protocols:
y 2% Basic
Source: 81.9% Basic
y 2% Screening
*Finished - Drinking water monitoring: 81.9%
screening, Confirmation, Comparison y 45% Screening
y 81.9% Basic
Stormwater
y 81.9% Screening
*TRE: Basic, Comparison y ASTM Basic

Recreational water y Basic Solid-Phase

y Basic
81.9% Basic, Comparison
y Comparison
Ecotoxicology y Confirmation
Basic, 2% Basic, 81.9% Basic, ISO, DIN y DIN
y International Standard ISO 11348-3
Emergency response
y Solid-Phase
45% screening, 2% screening, 81.9% screening y SOLO 2% Screening

Personal care/ Household chemicals y SOLO 45% Screening

(products, formulations and raw materials y SOLO 81.9% Screening
along with typical WWTP apps)
y Whole Effluent Toxicity (WET)
Basic, 45% Screening, 2% Basic, 2% Screening
2 Quality Assurance:
Food and beverage:
y Phenol Standard
* Water security monitoring program: 81.9%
Screening, Confirmation, Comparison y Zinc Sulfate Standard

Medical device
Basic

Drilling muds/ sediments

Solid Phase, Basic Solid Phase

9
 Microtox® - Acute Toxicity Manual Section 2

Methods in detail:
Basic: The procedure described is for freshwater
A 45% concentration sample test, used for samples, but it could also be used for marine
samples on unknown toxiocity. This test and estuarine samples except for using the
provides the most flexibility along with the appropriate diluent for sample preparation.
highest confidence level and test precision. The SPT procedure allows the test organisms
The test measures relative toxicity; to come in direct contact with the solid
producing data for calculating EC50 or IC50 sample in an aqueous suspension of the test
values or other ECxx or ICxx values. The test sample. Thus it is possible to detect toxicity
is recommended for: pure compounds, which is due to the insoluable solids that are
septage, wastewater influent, wastewater not in solution. Use for whole sediment or soil
treatment digester. (sediment or soil from which the liquid phase
has been removed).
81.9% Basic:
Designed for use with samples of low to Screening:
medium toxicity, such as post secondary For testing at a single concentration. No dose
treatment, wastewater effluent, stormwater response curve is generated and no relative
runoff, and pore water, can also be used for toxicity is determined.
toxicity baseline studies. Recommended for
2% Screening:
samples with expected EC50’s between 30%
Designed for use with samples of high toxicity.
and 100%.
This protocol is appropriate for samples such
2% Basic: as untreated industrial effluents, trucked
Designed for use with samples of high toxicity waste from industrial sources, and any other
(e.g. industrial waste streams) where an EC50 sample types of high toxicity. Microtox Acute
is required. Recommended for wastewater Reagent is used with this protocol. For the
influent streams and pure compounds. same test using Microtox SOLO reagent,
choose the SOLO 2% Screening Test.
ASTM Basic:
This method is currently an ASTM Standard 45% Screening:
Method (D5660). The “Standard Test Method Designed for use with samples of medium
for Assessing the Microbial Detoxification of toxicity. For the same test using Microtox
Chemically Contaminated Water and Soil SOLO reagent, choose the SOLO 45%
Using a Toxicity Test with a Luminescent Screening Test.
Marine Bacterium,” describes the effective 81.9% Screening:
use of Microtox to evaluate the ‘toxicity of Designed for use with samples of low toxicity.
wastewaters and aqueous extracts from This protocol is appropriate for the Microtox
contaminated soils and sediments prior Drinking Water Program. Microtox Acute
to and following biological treatment.’ Reagent is used with this protocol. For the
Basic Solid Phase: same test using Microtox SOLO reagent,
Designed for use with soil, sediment, and choose the 81.9% SOLO Screening.
other solid samples. The initial concentration Comparison:
is 99,000mg/L. The Basic Test should also Measures the relative acute toxicity of an
be used when the toxicity of the sample unknown sample against a control sample.
is unknown. The control sample can be either another
Solid Phase: sample of known toxicity or prepared from
This procedure allows the testing of a solid Microtox Reconstitution Solution. All samples,
sample using serial dilutions of the sample. control or unknown, must be osmotically
The procedure tests a sample at nine (9) adjusted. The Comparison Test is the best
dilutions with 9.9% (99.01 mg/L) being the protocol for testing low toxicity samples
highest concentration. when an ECxx value cannot be determined
using a basic test. The Comparison Test uses
10
Section 2 MODERN WATER
A Deepverge Company

multiple replicates of the sample at a single

concentration. The results are reported as
mean percent difference as compared to a
reference sample.
Recommended sample types are: Wastewater
treatment plant effluent, Storm water runoff,
Drinking water (high confidence testing),
Pore Water, Eluate.
Confirmation:
The best protocol for confirming the toxicity
of a sample. The sample is tested against
itself (with fresh preparation) to establish if
the first test result was accurate. This test is
primarily designed for use with the Microtox
Water Quality Monitoring Program.
Deutsches Insitut für Normung
(DIN) 38412 Teil Test:
Germany granted Microtox a DIN Standard
(DIN 38412 part 34) for the testing of water
samples as indicated in the ISO Standard
11348-3, allowing for fresh water, marine and
estuarine sample reporting.
International Standard Organisation
(ISO) 11348-3:
Microtox is an ISO Standard 11348-3: 1998.
The ISO Standard states that this method is
applicable to wastewater, fresh water
(surface and ground water), sea and brackish
water as well as elutes of sediment, pore
water, aqueous extracts and leachates.
Whole Effluent Toxicity (WET) Test:
This is not an EPA approved test.
This test is designed for use on wastewater
treatment effluent.
Phenol Standard:
Reference standard best chosen when sample
testing involves samples that contain organics.
Zinc Sulfate Standard:
Reference standard best chosen when
sample testing involves samples that contain
metals and other inorganics.

11
 Microtox® - Acute Toxicity Manual Section 3

Purpose of Standard
Testing a standard whose test results are well Zinc Sulfate Standard
characterized, confirms your understanding
of a test protocol and checks the performance EC50 15 minutes
of the complete test system (e.g. analyzer, ZnSO4.7H2O = 3 to 10 mg/L
Reagent, Diluent and Reconstitution Solution). Zn++ = 0.6 to 2.2 mg/L
The Basic Test Protocol (with two controls, 5. Weigh out ~ 50 mg (~ 0.050 g) of
eight sample dilutions in duplicate) is the ZnSO4.7H2O, add it to a 500 milliliter (mL)
best procedure for testing “Standards,” volumetric flask.
and provides a high level of precision Do not try to weigh out the exact
and flexibility. amount of sample; calculate sample
When testing “Standards” never accept concentration for the amount
an EC50 derived from extrapolated data, weighed out.
always retest using the appropriate 6. Add Diluent to the 500 mL mark on the
“Standard” dilutions. volumetric flask.
Phenol Standard 7. Seal the flask and mix well by inverting
the volumetric flask.
Phenol EC50 5 minutes = 13 to 26 mg/L
1. Weigh out ~50 mg (~0.050 g) of 8. Label the flask and store at normal
crystalline phenol, add it to a 500 refrigerator temperature (2-8°C).
milliliter (mL) amber volumetric flask. The Zinc sulfate standard will last for 3-4
Do not try to weigh out the exact amount of months when stored in this manner.
sample, calculate sample concentration for
the amount weighed out.
If an amber volumetric flask is not available,
cover the entire flask with aluminum foil to
protect the Phenol Standard from light.
2. Add Diluent to the 500 mL mark on the
volumetric flask.
3. Seal the flask, and mix well by inverting
the volumetric flask.
4. Label the flask and store at normal
refrigerator temperature (2-8°C).
The Phenol standard will last for 3-4 months
when stored in this manner.

12
Section 4 MODERN WATER
A Deepverge Company

Data Reduction Formulas

Gamma for the Microtox Acute where C is concentration, p is the number
of toxic molecules per target site, and k is a
Toxicity Basic Test Protocols composite factor relating to free energy
and volume changes of the reaction. This
Correction Factor relationship may be restated in the form
The Correction Factor (Rt) is the fraction of a linear equation for prediction of
obtained when the light output of the concentration from gamma values:
Control (Blank) remaining after time t is log C = b log Γ + log a
divided by the initial light output of the
untreated Control (Blank): This equation describes a line with a slope
of b and an intercept of log a, in which C
Rt = It ÷ Io represents concentration and Γ the
corresponding gamma. Logs of any
Gamma base may be used, as long as they are
The gamma value (Γ) is the ratio of the light used consistently.
lost at time t to the light remaining at time t
for a given sample concentration: The equations below utilise LOG
to denote logs (base 10)
Γt = ( RtIo - It ) / It = [( Rt x Io ) ÷ It ] - 1
Regression statistics of logC on log Γ are used
The second form of this equation stresses the
to estimate the concentration which would
fact that gamma is the ratio of light
give a nominal effect, for example an EC50.
expected for a nontoxic sample to that
In this example EC50 represents a gamma of
observed, minus (-) 1.
1. Since log 1 is zero, canceling the effect of
Gamma for the Microtox Acute the slope, the log C estimate is identical to
log a. Therefore, the EC50 is equal to the
Toxicity Solid- Phase Test Protocols antilog of the intercept (log a).
Concentrations which cause other percent
The gamma value (Γt) is the ratio of the light
effects (%E) can be estimated by substituting
output of the Control (Blank) at time t to the
the log of the equivalent gamma value in the
light output for a given sample concentration
regression equation.
at the same test time, minus (-) 1:
To convert %E to gamma, divide %E by 100
Γt = (Rtc / It) - 1
minus %E. Thus for an EC20:
Note that this is equivalent to making the
Γ = 20/(100-20) = 20/80 = .250 which is the
correction factor (Rt) equal to one in the
equivalent Γ. Percent E can be derived from
above equation for the Basic Test, which
gamma by:
means that no correction for reagent
pipetting errors is made. %E = [ Γ ÷ ( 1 + Γ )] ×100%

Regression Statistics Standard statistics are used to calculate

confidence intervals for this estimate. The
Rate theory for biological inhibition predicts specific equation for a 95% confidence
a simple mathematical relationship between factor in this case is:
the concentration of a toxic material and
CF=√[S2( 1 ÷ N )+(( log Γo-logΓa )2 ÷ Σγ2 )]
the response of a susceptible organism when
the response is measured in terms of
gamma values:
Γ= kCP
13
 Microtox® - Acute Toxicity Manual Section 4

In this equation Γo is the gamma greater, indicating that the variances

corresponding to the ECXX of interest, log Γa are not homogeneous, then a modified
is the mean of log (gamma) values, and N is t test is used to test for significance.
the number of data pairs used for regression.
The residual variance, S2, is calculated from 4. Percent difference, 95% confidence
the regression statistics as: limits, and a “t” value are calculated.
The calculated t statistic is compared
S2 = [ Σc2 - (Σcγ)2 ÷ ( Σγ2 )] ÷ ( N - 2 ) with look-up table values of t for 8 df
(homogeneous) or 4 df (not homogeneous)
The product of CF and t.05 (from a Student’s at different levels of significance from
t table) is added to and subtracted from log 60% and above.
Co to give the logs of the higher and lower
95% confidence limits. The antilogs are then 5. The results are given as percent
taken of log Co and its limits. difference ± the 95% confidence limits.
The confidence level of the difference
Comparison Test Data Reduction is also stated (60% - 99.9%). If the
confidence level of the difference is less
The Comparison Test compares a single than 60%, no confidence level is stated
dilution of a sample with a control, such as (indicated by “-”). If the difference is
diluent, or with another sample to determine stated as a negative number, it indicates
if there is a significant difference between that the sample was less toxic than the
the two samples. Five replications of each control (or the sample with which it was
control and sample are used. The statistical being compared.)
approach is conventional, and is referenced
in the bibliography. Terms:

This document discusses the approach in the Xc, Xs = Individual readings at time t for
specific terms of the test protocol. Note that control & sample
an equal number of replicates, fixed at 5
/ Xc, /Xs = Sums of the readings
each for sample and control, simplifies the / /
math involved over that of a general case. _ _
This approach assumes that the data are Xc, Xs = Means of the readings, = /X/n,
/
normally distributed, but is robust for where n = 5 = number of replications
moderate departures from normality. _ _
The fundamentals of the statistical 1 + ( 1 - Xs / Xc ) for zero time means =
calculations are outlined below: correction factor for timed sample data

1. Means, estimated variances, and / D 2, / D 2 = Sums of squares of differences

/ c / s
corresponding standard deviations are
calculated for the corrected readings from the mean, = / X2-( / X)2/n
/ /
for the control and the sample for each
exposure time. Sc2, Ss2 Estimated Variances, =/ D2/n-1
/

2. Variances are used to calculate an Sc, Ss Standard deviations, = √ S2

F statistic, which is compared with the
critical F distribution value for a two- F = Sc2 / Ss2 or F=Ss2 /Sc2
tailed, 95% confidence level. For 4x4 (larger variance is placed in numerator)
degrees of freedom (df) this value is 9.6. _ _
3. If the calculated F statistic is less than tobs = ( Xs - Xc / √ ( Dc2 + Ds2 )) (√20).
9.6, then a standard student’s t test is
used to test for significant differences [Note: √(nc ns /nc + ns) √(nc + ns - 2) = √ 20 for n=5]
of the means. If the F value is 9.6 or

14
Section 4 MODERN WATER
A Deepverge Company

Comparison Test Details:

1. Means are calculated for the zero time The t05 for 8 df from Table 1 is 2.306, thus
readings for control and sample. The the difference with limits is the difference
ratio of these means is used to calculate between the two means ± 2.306 Scs. The
a correction factor for the purpose of confidence level that the difference is real
normalizing the timed sample data. can be computed by comparing the
(See example below.) observed t, tobs, with the table values of t for
each probability level. If tobs is greater than
2. Means, variances, standard deviations,
the table value of t, then the difference is
and the F factor are calculated for each
significant at the table probability.
set of timed data. The result of the F test
for homogeneity will determine the Case of heterogeneous (unequal) variances.
method for testing for significance and In this case the treatment is the same as the
for calculating the 95% confidence limits. above case except that the degrees of
freedom are 5 -1 = 4, and the t05 for 4 df
3. The tobs is calculated from the above
is 2.776. The table 1 values for t correspond
given formula. The standard deviation
to 4 df.
used for calculating confidence limits is
the average of the two standard
deviations, Scs = Sc + Ss / 2
4. Case of homogeneous (equal) variances.
The degrees of freedom are
2(n -1) = 2 x 4 = 8.

Comparison Test Calculation Examples

The following raw data were obtained comparing a water sample against reference water in
a 5 minute test:

IO I5 I5 corrected
Control 97.12 103.91

Control 91.88 102.76

Control 94.75 102.93

Control 94.62 104.17

Control 93.88 103.24

Mean 94.45 103.40

Sample 90.88 93.78 95.32

Sample 90.16 94.52 96.07

Sample 95.05 94.66 96.21

Sample 94.83 94.86 96.42

Sample 93.58 95.21 96.77

Mean 92.90 94.61 96.16

15
 Microtox® - Acute Toxicity Manual Section 4

Comparison Test Details:

Correction factor, CF = 1 + ( 1 - 92.90/94.45) = The average of the two variances is:
1.01641. [indicates 1.01641% difference
between the zero time means] Corrected Scs2
sample I5 mean = CF x 94.61 = 96.16; individual (0.3773 + 0.2895) / 2
readings are corrected in the same way. 0.6668 / 2
The corrected values are used in subsequent 0.3334
calculations. For the five minute data
The standard deviation of the average is:
/ D 2 & / D 2 are 1.5091 and 1.158
c s
/ /
Scs
respectively as calculated by the formula in
√Scs2
Terms above. The variances are thus these
numbers divided by 4 (df), or 0.3773 for the √0.3334
controls and 0.2895 for the samples. F is the 0.5774
larger of these variances divided by the This value is used to calculate the 95%
smaller; 0.3773 / 0.2895 = 1.303. The calculated confidence limits. The difference between
F is smaller than the critical F value of 9.6, so means is:
these variances may be considered equal
and the standard t test may be used. 103.4 - 96.16 = 7.24
The limits are ± 2.306(.5774), yielding 7.24 ±
From the formula for t: 1.33. These are converted to percent
tobs differences by:
( Xs - Xc / √ ( Dc2 + Ds2)) (√20) 7.24 / 103.4 x 100 = 7.00% and
( (103.4 - 96.16) / √(1.5091 + 1.158)) (√20) 1.33 / 103.4 x 100 = 1.29%
(7.24 / √2.6671)(√20) Thus the results for 5 minute data are
19.83 reported as:
From the look-up table, Table 1, it can be seen % Difference 7.0 ± 1.3
that the calculated t is greater than the 99.9% Significance level 99.9%
look-up value for 8 df of 5.041 indicating that
the difference between means is significant
with a probability greater than 99.9%.

Table 1: Look-up table for standard and modified t tests*

Significance Level % t for 8 df (standard) t for 4 df (modified)

60 0.889 0.941

80 1.397 1.533

90 1.860 2.123

95** 2.306** 2.776**

97.5 2.752 3.495

99 3.355 4.604

99.5 3.832 5.598

99.9 5.041 8.610

** values used for the 95% confidence limits calculations

16
Section 5 MODERN WATER
A Deepverge Company

Special Sample Handling

Difficult Samples Made Easy 4. Samples that have an extremely steep
dose-response curve
In several situations, samples may require Some test samples show a dose-response
special handling. curve so steep that the light level drops
from very high to very low from one
1. Samples that have known sample concentration to the next,
special requirements providing too few useful data points for
Some samples are already saline, requiring calculation of ECXX.
omission of the osmotic adjustment step; Complete light loss may occur at all
others are colored, others have unusual concentrations in an Abbreviated Basic
physical characteristics. Such samples Test Protocol. After a second test with a
are characteristic of certain industries primary dilution, all concentrations may
and activities and various show no light loss at all. A trial-and-error
manufacturing situations. search for the narrow range in which the
This manual contains a number of change occurs can be time-consuming.
variations on the fundamental Microtox This sort of data has a dose-response
Acute Test protocols designed specifically curve so steep at times that it cannot be
for dealing with samples known to require plotted by the system.
special handling. Examine these carefully, Use narrower (1:1.5) serial dilutions.
and select one that serves your
particular need. 5. Samples that have a negative slope
Some test samples show a negative slope
2. Samples that are exceptionally toxic (reverse dose- response curve). The light
Some samples are so bioreactive that level appears to increase with increase in
the tests you normally run show complete sample concentration. This condition is
light loss at all concentrations. Retest the caused by one of three mistakes:
sample, using the Basic Test Protocol and
make a “primary dilution” of the sample. 1. The sample dilutions were made in
With adequate dilution, the ECXX of any reverse sequence.
highly toxic sample can be found. 2. The It light readings were read in
It is desirable to use the Basic Test Protocol reverse sequence.
when testing a sample of completely 3. The most common cause is
unknown toxicity. background noise (pipetting error)
which by chance forms a
3. Samples that are of exceptionally negative slope.
low toxicity
Background noise can be more noticeable
The toxicity of some samples is extremely
with low toxicity samples and the 81.9%
low, but important to know. When the
Basic Test protocol.
toxicity levels are very low, they may be
Run the test protocol used, in duplicate,
confused with the background “noise”
using two controls with higher sample
produced by irregularities in pipetting.
concentration if possible.
Run an 81.9% Basic Test protocol, in
During the test, be very careful in
duplicate, using two controls in each set.
pipetting, making the sample dilutions
When the 81.9% Basic Test protocol does
and reading the light levels.
not provide ECxx, run a Comparison Test.

17
 Microtox® - Acute Toxicity Manual Section 5

6. Samples that have a flat

dose-response curve
Some samples show dose-response curves
so flat (at 1:2 dilution ratio) that the data
points may be confused with background
(pipetting) noise, and calculation of ECXX
may require extrapolation to a distant
point, widening the confidence range
beyond cceptable range limits.
Rerun the test, using the wide (1:10)
serial dilutions.

7. The ECXX is determined by extrapolation

Some samples are so bioreactive, or so
slightly bioreactive, at the range of
concentrations tested that all of the
data points fall on one side of the ECXX,
which can only be determined by
extrapolation. This is always less desirable
than an ECXX bracketed by data points.
If the gammas are all larger than that
corresponding to the ECXX of interest, all
gammas greater than 1.0 for EC50, retest
the sample using a primary dilution
testing lower concentrations. The problem
with extrapolation to determine ECXX is
that the sample may be nonlinear at
lower or higher concentrations.
If the gammas are all smaller than that
corresponding to ECXX (gammas less
than 1.0 for EC50) retest using a more
concentrated sample.

18
Section 6 MODERN WATER
A Deepverge Company

FAQs
These common Questions & Answers provide insight into the decisions and practices that
produce high quality data.

Q. How do I know which test to run? because the few sample dilutions used
were in the wrong range.
A. The Applications and methods section of
the manual has detailed descriptions of For unknown, or highly variable samples,
each protocol, choose the test that is the Basic Test Protocol is appropriate. This
appropriate for your application. protocol is much more likely to bracket
the ECXX of interest.
Q. Which test should be used for samples
that show no toxicity at concentrations of Q. Examples of tests in this manual use 1:2
45% and lower. serial dilutions to prepare the serial
dilutions of the sample.
A. Use the 81.9% Basic Test protocol or the
Wouldn’t a higher ratio provide a greater
Comparison Test.
range of concentrations?
Q. Duplicate tests. What are they, and when
A. Yes, but it reduces the resolution of the
would they be used?
test. The 1:2 dilutions are a good standard
A. The use of duplicates provides additional compromise between range and resolution,
data to check the quality of the test and work well for most samples.
(operator pipetting), and provides a larger
context in which to evaluate the data. Q. How long can the frozen Microtox Acute
Test Reagent be stored before it is used?
Q. How many controls should be used?
A. The shelf life of the reagent, from date
A. The number of controls can range from of manufacture, is at least two years
one to three. The highest confidence in conditions. See the discussion of Microtox
results would come from running three Acute Test Reagent in the FUNDAMENTALS
controls. Multiple controls provide another section of this manual. The expiration date
measure of pipetting consistency. of the reagent is printed on the label on
Q. How many sample dilutions should be the reagent vial. Make a point of using up
used? the reagent with earlier expiration dates
before opening new boxes of reagent with
A. When dealing with well-characterized later expiration dates. First in/first out.
samples, four sample dilutions that have
at least one dilution on either side of the Q. How can I tell if the fresh reagent always
ECXX are enough. In that case, you will produces the same amount of light?
typically be testing samples from the A. The reagent does not always produce the
same source on a regular basis, looking same amount of light, and the question is
for significant changes in toxicity levels. of little practical importance. The system
Tests using more dilutions are important if measures the percent of light lost in
you are testing samples with which you response to change in toxicity, which is
have no previous experience, and you consistent, no matter what the original
don’t know what to expect. light level. Light levels may be different for
several reasons:
For instance, if you run a test with just a
few dilutions of a very toxic sample, every a) The light output of different batches of
concentration may cause complete loss of reagent varies.
light output from the reagent. This indicates
b) Poor reconstitution technique reduces
that the sample has detectable toxicity,
the potential light output of the reagent.
but does not allow calculation of ECXX,

19
 Microtox® - Acute Toxicity Manual Section 6

c) The light output will change with the To establish baseline information, you may
age (shelf life) of the reagent. also elect to test each new lot of Microtox
Acute Test Reagent with a standard such
These variations are unimportant because
as Phenol or Zinc sulfate over a period of
the gradual light loss or gain during a
hours after reconstitution, so you know what
test is measured, and accounted for in
the normal response curve is over time.
data reduction calculations of the
Test Procedures. Once again, the decision on how long the
reconstituted reagent may be used is
In general, light output is not the primary
based on what information the test is
consideration in evaluating the quality of
intended to produce.
the reagent.
Q. Some of the protocols call for transfer of
More important is the response of the
only 10 µl of reagent mixture to the test
reagent to various toxic materials.
cuvettes. Can’t we increase that to 20 µl
Q. How long can the reagent be used or 30 µl, as long as we do it every time?
after reconstitution?
A. No, don’t increase the amount of reagent.
A. For at least three hours after reconstitution,
the organisms respond consistently to a Q. How can we measure and improve
wide spectrum of toxic materials. After that pipetting skills?
time, the response may begin to change. A. The Microtox Acute Test requires transfer
The reagent does not just become generally of small amounts of liquid, as little as
“less sensitive.” Its response to different 10 µL at a time, with an assortment of
chemicals may either increase or decrease, micropipettors. Pipetting is an ordinary
changing the response spectrum. laboratory skill and usually improves
The useful life of the reagent after with repetition.
reconstitution is nominally three hours. To check your reagent pipetting
Many laboratories reconstitute a vial of consistency:
reagent for use all morning or afternoon.
Using an analytical balance, transfer 10 µL
The conditions under which you use
of water into a tared cuvette and weigh
Microtox Acute Test Reagent beyond the
the amount of water pipetted. Repeat this
initial three hours will depend on your need
10 times and calculate the mean.
for precision and verifiability of results.
Q. How can we calibrate a pipettor?
For instance, if you intend to use the data
for routine monitoring, and data quality A. Using an analytical balance, weigh the
requirements may not be critical, the use amount of water transfered into a tared
of reagent over several hours may be cuvette. Repeat this 10 times and
perfectly acceptable. If unusual or calculate the mean.
suspicious results are seen, these results Q. Sometimes several or all of the sample
could be confirmed with tests using a cuvettes produce negative gammas. Does
“fresh” bottle of reagent. that always indicate a stimulatory effect
Well-characterized samples can be tested of the sample?
repeatedly with reagent of increasing A. Not always. The effect may also be
age, so that the “normal” response of the produced by pipetting error.
reagent at any age is well known. The
Phenol Standard is useful for this. It is also
helpful to run several tests on one or more
well characterized stable samples of
known toxicity.

20
 MODERN WATER
A Deepverge Company

To find out how we can help you please contact us on:

US: +1 302 669 6900 UK: + 44 1904 404036 IR: +353 1 556 8250

Modern Water Inc Modern Water Modern Water

15 Read’s Way, Suite 100
Monitoring Ltd Monitoring Ltd
New Castle Modern Water UK, Modern Water European
DE 19720 York Biotech Campus, Warehouse, Glanaco, IDA
United States Sand Hutton, York YO41 1LZ, Industrial Estate, Charleville,
United Kingdom Co. Cork. Ireland P56 T275

China: +86 (0) 21 6230 6163

Modern Water Technology

(Shanghai) Co. Ltd
#1702 Xinyin Building, No. 888, Yishan Road
Xuhui District, Shanghai 200233, P.R. of China [email protected]
21
www.modernwater.com