How to optimize the drop plate method for enumerating bacteria

https://doi.org/10.1016/S0167-7012(00)00241-4Get rights and content

Abstract

The drop plate (DP) method can be used to determine the number of viable suspended bacteria in a known beaker volume. The drop plate method has some advantages over the spread plate (SP) method. Less time and effort are required to dispense the drops onto an agar plate than to spread an equivalent total sample volume into the agar. By distributing the sample in drops, colony counting can be done faster and perhaps more accurately. Even though it has been present in the laboratory for many years, the drop plate method has not been standardized. Some technicians use 10-fold dilutions, others use twofold. Some technicians plate a total volume of 0.1 ml, others plate 0.2 ml. The optimal combination of such factors would be useful to know when performing the drop plate method.

This investigation was conducted to determine (i) the standard deviation of the bacterial density estimate, (ii) the cost of performing the drop plate procedure, (iii) the optimal drop plate design, and (iv) the advantages of the drop plate method in comparison to the standard spread plate method. The optimal design is the combination of factor settings that achieves the smallest standard deviation for a fixed cost. Computer simulation techniques and regression analysis were used to express the standard deviation as a function of the beaker volume, dilution factor, and volume plated. The standard deviation expression is also applicable to the spread plate method.

Introduction

The drop plate (DP) method exhibits many positive characteristics. The plating and counting procedures require less labor than alternative methods. The plating and counting steps are very convenient and manageable. On appropriately dried plates, the drops will absorb quickly into the agar. By distributing the sample in drops, colony counting can be done faster and perhaps more accurately. The drop plate method expends relatively few supplies. A bibliographic-database search and a worldwide web search showed that the drop plate method is being used in numerous laboratories across the world. In spite of its widespread use, the DP method has not been standardized.

Early advancements in the development of the drop plate method are accredited to several authors. These authors described, in brief form, the adaptation of dropping pipettes to the technique of plate counts of bacteria, particularly Wilson (1922), Aitken et al. (1936), Kenny et al. (1937), von Haebler and Miles (1938), Miles et al. (1938), Snyder (1947), Reed and Reed (1948), and Badger and Pankhurst (1960). The Miles et al. and Snyder papers include a statistical analysis of the accuracy of the method. In particular, Miles et al. derived the variance of the plate counts using binomial and Poisson distributions where the plate counts were averaged over all drops and plates. Badger et al. tested the effects of the use of different pipettes, of variations between drops from the same pipette, of variations between successive fillings of a pipette from the same dilution, and of variations between plates. The results of these tests show that there is no significant difference between pipettes, drops, or plates.

The plating process distinguishes the drop plate (DP) method from alternative methods. The most popular alternative is the spread plate (SP) method. In the SP method, 0.1 ml of a liquid sample is inoculated onto an agar plate. The liquid sample is spread into the agar with a flame-sterilized hockey stick, immobilizing the cells on the surface of the agar. The colony-forming units (CFUs) are counted after an appropriate incubation period. For the DP method, however, the sample volume is dispensed on the agar plate in a fixed number of separated, small drops. After incubation of the plates, the colonies within the drops are counted and the counts are scaled up to estimate the total number of CFUs in the initial beaker volume. Because counting is confined to the drops, the DP method is not recommended for organisms that display a swarming type of motility; e.g., Proteus mirabilis, P. vulgaris, and Vibrio parahaemolyticus.

Accurate and precise measurement of the drop volume is absolutely necessary to the DP method. Donald (1915) was the first to describe a method for the precise measurement of fluid volume by means of drops. Fildes and Smart (1926) expanded the procedure and developed methods of preparing and calibrating the pipette. Today, an electronic pipetter, costing less than $400 US, possesses the qualities of high accuracy and precision.

The DP method is a mixture of microbiological components and design components. The microbiological factors are fixed by the purpose of the experiment. They include the bacterial species, strain, and growth conditions (e.g., media, agar, temperature, time). This paper will focus only on the design factors (i) beaker volume, (ii) dilution factor, and (iii) volume plated. Throughout this paper, any specific combination of levels of these three factors will be called a design case. Beaker volume is synonymous with initial culture volume. The bacteria in the beaker may have originated in a sample from the environment or experimental apparatus. The sample could be a volume of liquid from a laboratory chemostat, recreational water, or drinking water. If the sample were a semi-solid, such as sediment, soil, or food, it would be blended in with a liquid, thereby creating the beaker volume suspension of disaggregated bacteria. In biofilm studies, it is common practice to remove the biofilm from a known surface area and disaggregate the bacteria in the beaker. Note that disaggregation methods are outside the scope of this paper. Our analyses assume that the bacteria have been properly disaggregated and are randomly mixed in the beaker.

To obtain distinct, non-overlapping colonies on the agar plate, the sample to be counted must almost always be diluted. Since the technician has only a rough guess of the viable count ahead of time, it is usually necessary to make more than one dilution. The dilution factor is a number defining the level of dilution. A larger dilution factor indicates a higher multiplicative fold dilution. For example, a dilution factor of 10 specifies 10-fold dilutions of the sample and a dilution factor of two specifies twofold dilutions. This factor partially governs the length of the dilution series. Density estimates based on 10-fold dilutions will usually involve fewer total dilutions than if based on twofold dilutions. Twofold dilutions, however, usually improve the precision of the density estimate.

Many different designs exist and have been implemented in laboratories. For example, the volume plated has varied from 0.1 ml (10 drops of 10-μl volume; Zelver et al., 1999) to 0.12 ml (six drops of 20-μl volume; Miles et al., 1938) to 0.15 ml (six drops of 25-μl volume; Reed and Reed, 1948). To apply the DP method, laboratory technicians must choose the number of drops that make up the volume plated. However, the number of drops does not directly affect the standard deviation of the density estimate. The number of drops is incorporated into the calculation only because the volume plated equals the number of drops times the drop volume. The calculations are the same for 10 drops of 20 μl as for 20 drops of 10 μl, which are the same as for any combination of number of drops and drop size, which when multiplied together equals a volume plated of 0.2 ml.

The outcome (or result) of the DP method is an estimate of the density of microorganisms with units CFUs per beaker volume. The precision of the density estimate is indicated by its standard deviation (SD). The plating of a larger volume leads to more information about the true density of microorganisms. In turn, more prior information about the true density leads to less variability in the density estimate; i.e. a smaller SD. In theory, if the entire beaker volume is plated and counted, the SD would be zero. Of course, plating and counting the entire beaker volume is not realistic in terms of cost. Therefore, a compromise between precision and cost must be made.

The goals of this paper are to determine (i) the SD of the bacterial density estimate, (ii) the cost of performing the DP procedure, (iii) the optimal DP design, and (iv) the advantages of the DP method in comparison to the SP method.

Section snippets

Drop plate method

In this study, the laboratory experiments were run according to the following protocol. Pseudomonas aeruginosa (ERC-1) was used for all experiments. Using proper aseptic technique, an isolated colony was inoculated into a flask containing 100 ml of sterile Tryptic Soy Broth (Difco) at 300 mg/l. The flask was allowed to incubate in an orbital incubator at 35°C for a maximum of 24 h. The viable cell density was approximately 108 CFU/ml.

All serial dilutions were performed using sterile buffered

Results

The time estimates found in our laboratory were averaged and used to construct our time function. The estimated time involved in completing one dilution in the dilution series was A=22.8 s. The estimated time involved in plating one dilution was B=44.4 s. Eq. (7) specifies the function used as the time constraint in this study.T(k)=67.2k+44.4,fork=0,1,222.8k+4·44.4,fork=3,4,5.Eq. (8) is the multiple regression model for LSD, when the true density is λ1, where λ1=104, λ2=3.16×105, λ3=106, λ4

Discussion

The recommended designs for each λ (Table 1) may not be suitable for every technician in every laboratory. We suggest the following steps to determine an efficient DP design. First, estimate the time required to complete one dilution (A) and the time required to plate one dilution (B) (see Eq. (6)). Given the time estimates, construct a time function similar to Eq. (7) (utilizing Eq. (5) for k). Second, determine the DP design cases of interest by specifying the factor levels of BV, DF, and VP.

Acknowledgements

This work was funded by the National Science Foundation Engineering Research Centers Program (Cooperative Agreement #EEC-8907039) and the Environmental Protection Agency (EPA) under Contract #68-W-99-015 with Montana State University-Bozeman. This paper does not necessarily reflect the views of the EPA.

References (16)

  • R.S. Aitken et al.

    A case of botulism

    Lancet

    (1936)
  • N. Zelver et al.

    Methods for measuring antimicrobial effects on biofilm bacteria: from laboratory to field

  • Anonymous

    Rainin Pipetting Solutions, 1998

    (1997)
  • E.H.M. Badger et al.

    Experiments on the accuracy of surface drop bacterial counts

    J. Appl. Bacteriol.

    (1960)
  • R. Donald

    A method of drop-measuring liquids and suspensions

    Lancet

    (1915)
  • K. Donegan et al.

    Evaluation of methods for sampling, recovery, and enumeration of bacteria applied to the phylloplane

    Appl. Environ. Microbiol.

    (1991)
  • P. Fildes et al.

    Volumetric measurement by drops in bacteriological technique

    Br. J. Exp. Pathol.

    (1926)
  • R.A. Fisher et al.

    The accuracy of the plating method of estimating the density of bacterial populations

    Ann. Appl. Biol.

    (1922)
There are more references available in the full text version of this article.

Cited by (683)

  • Tailoring AA6063 for improving antibacterial properties

    2024, Applied Surface Science Advances
View all citing articles on Scopus
View full text