I actually have algae growing in water in a container. I was thinking if it was possible to filter the water so that all the small cells will be filtered out and only the bigger ones will remain to reproduce cells that tend to be bigger. Is that possible? What type of filter should I use?

  • $\begingroup$ A formal showing of your 'selection pressure' might also require proof that your initial algae population consists of only 1 type of algae. $\endgroup$
    – PlaysDice
    Commented Apr 17, 2014 at 0:26
  • 1
    $\begingroup$ Wouldn't this actually be artificial selection? $\endgroup$
    – evamvid
    Commented Apr 17, 2014 at 1:59
  • $\begingroup$ @evamvid which is a form of evolution. $\endgroup$
    – John
    Commented May 27, 2023 at 23:55

2 Answers 2


Short answer

Yes that would work in the condition that the trait you select for (size) is heritable.

Long answer

The kind of selection you would apply is called truncated selection because you fix a limit in size (depends on your filter) under which individuals do not survive and above which individuals survive and reproduce equally.

The response to selection ($R$) is given by the breeder's equation:

$R = h_N S$

, where $S$ is the selection you apply. In your case, $S$ is the difference between the mean size of the parent generations and the mean size of those you allow to reproduce in the parent generations. $h_N$ is the heritability (more info in additional info).

So, if you measure the mean size of your cells, filter them, measure their size again (or infer their size based on previous measurement and the size of you filter) and calculate the difference between the two measurements you get $S$. Then you let your filtered individuals to reproduce and you measure the size of their offspring which is $R$. By dividing $R$ by $S$ you get the heritability $h_N$. If you do not observe any difference in size between your two generations it means that the heritability is 0 ($\space h_N = 0\space$). Of course, you'll need a population that is composed of enough cells in order to have a good estimate.

Additional info

To be a bit more accurate $h_N$ is the heritability in the narrow sense. It represents the part of variance in the population that has to be accounted to additive genetic variance. The word additive genetic variance might seem obscure. In other words

$$h_N = \frac{V_{Ga}}{V_{P}} = \frac{V_{Ga}}{V_{G}+V_{E}} = \frac{V_{Ga}}{V_{Ga} + V_{Gd}+V_{E}}$$

, where $V_{GA}$ is the additive genetic variance, $V_{GD}$ is the dominance genetic variance, $V_E$ is the variance due to environmental variance and $V_P$ is the total variance (phenotypic variance)

Hope that helps. Do not hesitate to ask further question if something is unclear or if I missed one of your point.



The filter you need should allow you to discriminate the cells based on their size. Therefore, it should be relatively accurate. It is common to classify the phytoplankton based on their size, see this. You should try different filters and see what filter allows a good discrimination.


I assumed your cells were all of the same species. Is it the case? If you just sampled them from a lake and let them grow in a container then you very probably have various species. Therefore, you would select (based on their size) for species with your filter rather than selecting individuals within a species (based on their size again). It is kind of the same process anyway and the breeder's equation still holds.

  • $\begingroup$ thanks for the very helpful answer. can you tell me what filter i can use? $\endgroup$ Commented Apr 16, 2014 at 20:25
  • $\begingroup$ @SumitSingh See my updated answer. $\endgroup$
    – Remi.b
    Commented Apr 16, 2014 at 21:16
  • $\begingroup$ I just saw your edit re single species after I had commented the question. OP sounds like a fun science project, if it can be performed accurately! $\endgroup$
    – PlaysDice
    Commented Apr 17, 2014 at 0:39

I think one would not have a lot of success using a filter to do this, rather using a centrifugal gradient and comparing the fractions to a standard I think would be preferable. Nitrocellulose filters are made with precision but not very astounding accuracy which is the main reason they are commonly used at set sizes (like 0.2 and 0.4 microns) and not often in between. Good question though


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .