The Most Probable Number Method
The Most Probable Number Dilution-Culture Method (MPN method) measures the number of viable phytoplankton cells in a sample, via their ability to reproduce. It is a formal mathematical calculation based on binary scoring data from a set of dilutions, and replicates from a sample. In a ballast water management application, the binary scoring is of reproduction or no reproduction of phytoplankton, in dilutions and replicates of a ballast water sample.
Here’s how the MPN method is performed to measure biological efficacy in a ballast water management application:
- A ballast water sample is serially diluted
- The dilutions are incubated at favorable light and temperature levels together with nutrients and purified water constituents from ambient water
- Over time, dilutions are monitored for chlorophyll fluorescence with a standard laboratory fluorometer. If one or more viable cells are present in any dilution, they will reproduce and increase the chlorophyll fluorescence
- Dilutions are simply scored for growth or no growth, based on changes in chlorophyll fluorescence over the incubation period
- Scores at each dilution are used in the MPN calculation to determine how many viable cells were present in the original sample
Cullen, J.J. and H.L. MacIntyre (2015). On the use of the serial dilution culture method to enumerate viable phytoplankton in natural communities of plankton subjected to ballast water treatment. J. Appl. Phycol. DOI 10.1007/s10811-015-0601-x. http://creativecommons.org/licenses/by/4.0/
The 10 – 50 μm Size Class
The 10 – 50 μm size class in ballast water samples can contain both phytoplankton and heterotrophs. The MPN method is well-suited to phytoplankton, as they can be cultured reliably and repeatably in ballast water samples. Heterotrophs in the same size class, however, are more difficult to culture. As a result, an alternative 2-part method was developed and included in the aforementioned application submitted to the USCG. The alternative method utilizes MPN to enumerate viable phytoplankton, and microscopic evaluation of motility to enumerate viable heterotrophs.
Relative Accuracy & Repeatability of Methods
Substantial empirical data from tests using both the MPN and vital stain methods provides an important indication of the relative accuracy for a given test or organism species. The ideal situation is that a method is able to achieve 100% accuracy in classifying the organism correctly, but often methods do not achieve this standard. In some instances, a method can completely fail to accurately classify a given test or organism species and thereby diminish a methods ability to assess compliance with either USCG or IMO BWM Convention standards.
In the images below, the center of the target represents 100% accuracy in a given test or organisms species. 0% represents complete failure in a method’s accuracy for a given test or organism species.
The data for the MPN diagram presents a very conservative picture as the accuracy is presented and calculated on a species basis versus a number of organisms basis (“abundance basis”). The abundance basis calculation is a more relevant value as it is consistent with the USCG’s Ballast Water Discharge Standard (“number of organisms/mL” and not “number of species/mL”). If the data were to be presented and calculated on an abundance basis, all the data points would be at the center of the diagram as a much higher percentage of organisms are shown to grow out on an abundance basis versus a species basis.
A Simplified Example
Think of it as a series of dilutions to determine the point of disappearance. From the point of disappearance, the number of viable cells in the original sample can be back-calculated. If cells are detected in the 10-fold dilution but are not detected in the 100-fold dilution, the original sample must have had between 10 and 100 cells. Through the use of replicates and the mathematics underlying the MPN calculation, a much more precise result can be determined.
- A sample is diluted three times in 10-fold steps
- From each dilution, five replicates are incubated
- During incubation, viable cells will reproduce, chlorophyll levels will increase, and fluorescence signals will increase
- Any replicate with at least one viable cell at the beginning of the incubation, will show an increase in fluorescence and will be scored as positive for growth
- Any replicate with no viable cells at the beginning of the incubation, will not show an increase in fluorescence, and will be scored as negative for growth
- The dilutions, the numbers of replicates, and the numbers of positive scores at each dilution, are used to make the MPN calculation
Mixed Assemblages and Competition
Ballast water samples can contain a mixture of phytoplankton species. They may grow at different rates, and a fast-growing species could outcompete a slow-growing species. There is a misconception that this makes the assay blind to slow-growing species. Fortunately, this is not the case.
Viable cells of all species are distributed through the matrix of dilutions and replicates. The accuracy of the assay relies on detection of growth. In a competitive situation between two species in the same replicate, as long as one of them grows to detection, the correct scoring result will occur. At the extreme dilution where replicates contain either one or zero viable cells, there is no competition. Thus the assay is not blind to certain species due to competition.
This MPN method is applied to phytoplankton in the 10 – 50 μm size class. Phytoplankton are autotrophs, or more specifically phototrophs. These organisms use light energy to form nutritional organic substances from inorganic substances in their environment. With light and nutrients, they are relatively easy to grow or culture in the lab, and are very amenable to measurement in the MPN method.
There is a common misconception that this method cannot be used for ballast water samples because it is not known how to properly culture the vast number of phytoplankton species in nature. This misconception is not supported in the literature, where a thorough review failed to find evidence that “unculturability” was some inherent property common to many phytoplankton (Cullen and MacIntyre, 2015). The misconception is probably carried over from other applications where phycologists strive to isolate and maintain distinct species in perpetuity in a defined media (a different definition of culturing). Those applications are difficult because the culturist must define their media and cannot rely on the suite of dissolved compounds that were in the water when the isolate was collected.
Culturing phytoplankton in ballast water is highly successful because sterilized local water is used to make the dilutions, so the phytoplankton have access to a non-limiting supply of the same dissolved compounds that they were adapted to. Only a few generations of growth are required to accurately detect growth, so these compounds are not depleted during the assay (which is typically only two weeks). Together with some added nutrients to provide ample building blocks for new cells, light energy to drive growth via photosynthesis, and a favorable temperature environment (temperatures near ambient to prevent shock, but slightly elevated to increase growth rates and shorten detection times), culturing of phytoplankton using the MPN method has proven to be very successful by those who practice it.
Measuring Success in Culturing
To measure the success of culturing a mixed community sample, painstaking microscopy is required:
- Untreated control samples are examined before and after the incubation period, and the species observed are recorded
- Growth is assigned to species that are observed in the post-incubation samples
- The growth verification method is instructive but by no means definitive
This is prone to issues of microscopic counting resolution, where low abundance species are difficult to detect in the presence of high abundance species. The MPN method strives to promote growth, leading to conditions that make verification of growth of slow-growing species difficult. The accuracy of the MPN method itself is not impacted by competition or by slow-growers, but accurate verification of growth is difficult. Another approach is to calculate growth based on demonstrated physiological ability of species to grow in the culture conditions. This data is available at some ballast water management system test facilities that maintain databases of species that have been observed to grow in their MPN assays.
To support the development of the MPN method, a number of experiments were done by ballast water treatment system test facilities to determine the success of culturing. After identifying optimum growth conditions, growth success was determined in individual MPN trials, based on the species present and their physiological ability to grow. Growth success ranged from 92 – 100%. Recognizing that abundance weighting is more meaningful for the discharge standard (meaning that growth success is not calculated on a per species basis, but on an abundance-weighted per species basis), growth success ranged from 96 – 100%.
For those that have tried to culture phytoplankton in the context of the MPN method applied to ballast water samples, very high growth success rates can be achieved, proving the assay is reliable and repeatable. This work was commissioned by a consortium of UV-based ballast water management system suppliers, and was submitted to the USCG as part of applications for approval of the use of the MPN method to determine biological efficacy of UV-based systems.
Cullen, J.J. and H.L. MacIntyre (2015). On the use of the serial dilution culture method to enumerate viable phytoplankton in natural communities of plankton subjected to ballast water treatment. J. Appl. Phycol. DOI 10.1007/s10811-015-0601-x.
1. MPN: data extracted from ETV Method Development Experiments and land-based test reports (DHI-Denmark).
2. MPN Method Analysis: Cullen, J.J. and H.L. MacIntyre (2015). On the use of the serial dilution culture method to enumerate viable phytoplankton in natural communities of plankton subjected to ballast water treatment. J. Appl. Phycol. DOI 10.1007/s10811-015-0601-x.
3. MacIntyre 2016: Stain data extracted from MacIntyre H.L., and Cullen, J.J. 2016. Classification of phytoplankton cells as live or dead using the 2 vital stains fluorescein diacetate and 5-chloromethylfluorescein diacetate (FDA and CMFDA). J. Phycol. DOI 10.1111/jpy.12415-15-154.
4. NRL 2010: Stain data extracted from Multi-site validation of a method to determine viability of organisms ≥ 10 µm and < 50 µm in ships’ ballast water using two vital fluorescent stains (NRL, 2010).