Plate Count Anomaly


As with most of the experiments in this course, you will be handling a variety of undomesticated organisms of unknown identity or pathogenicity. Handle all cultures with respect and using standard microbiological procedures.


Since the early days of microbiology, it has been known that cell counts of environmental samples obtained by cultivation (plate counts or MPN) are much lower, by many orders of magnitude, than direct microscopic cell counts. In this lab, the usual ratio is about 10^-6 (i.e. one cell in a million seen grows up as a colony), ranging from 10^-2 (1%) to <10^-8. Some of this discrepency is attributable to differing requirements of organisms, e.g. aerobic plate counts on rich media will miss all obligate anaerobes and autotrophs. In other cases, organisms are known to enter a noncultivatable resting state, and many organisms rely on each other for any of a variety of reasons and cannot be cultivated in isolation. Imagine mixing all of the nutritional requirements of a rabbit (carrots, water, air) in a huge fermentor, innoculating with a big chunk of forest, and hoping to culture rabbits! In any case, it is clear that what is readily cultivable is a miniscule fraction of what's in a sample & the organisms that do grow are almost by definition unlikely to represent the population as a whole.


  • pond water (from your Winogradsky column)
  • Petrov-Hauser counters
  • cover slips
  • Phase-contrast microscope
  • P20, P200 pipet guns and tips
  • 0.1 ml sterile water blanks
  • hockey sticks, EtOH
  • LB plates


  1. Place a small drop of pond water onto a Petrov-Hauser slide and cover with a coverslip.
    hemocytometer loading
  2. Examine microscopically using phase-contrast. Locate the center of the counting grid (with the smallest squares) and count the number of bacterial cells in 25-50 of the grid squares.
    View of the grid

    Here is what the Petrov-Hauser counting grid looks like:grid
    The sides of the grid have shoulders to hold the coverslip at a defined height of 20um.

  3. Add 10 and 1 microliter of the same water sample to each of two 0.1ml water blanks and mix.
  4. Spread these two dilutions and 100 microliter of undiluted sample onto separate LB plates with a flamed hockey stick. Incubate 2-7 days at 30C.

  5. Count the colonies and determine the 'cultivable cell count'. Pick the plate with 30-300 colonies, if you can.


Calculate the concentration of observable cells in the pond water based on the volume of water counted and the total cell count. The dimensions of each of the small grid squares in the Petrov-Hauser counter is 50um x 50um by 20um deep, or 5 x 10^-8 ml. So, take the average number of cells observed per small grid square (it's OK if this is less than 1), divided by 5 x 10^8 - the result is the number of cells obsvered per ml.

Also calculate the number of colony-forming-units (cultivable cells) per ml in the pondwater. This is just the number of colonies on a plate divided by the volume plated (in ml) - for example, 125 colonies from 0.01ml = 12500 cfu/ml.

Now calculate the fraction of observable cells that yielded colonies under these cultivation conditions (colonies per ml divided by cells per ml X 100%). Typical results are on the order of 0.0001%, ranging from 0 (no colonies) up to sometimes as high as 5%.