You can do per dip density. Use 300 ml bowls/dippers/ ladles for the purpose. Take 5 samples from the edges of the pool if the surface area is less than 1 sq mt. You can increase the number of samples as the area increases. Take the average. You can count the immature stages as I+II instar, III+IV instar and pupa.

Thank you Sumodan for your contribution.

I fully endorse the views of Dr. P. K. Sumodan

I agree with just little modification, you have to take samples randomly with different distance from the edges and don't stock only to them, because there is some factors affecting the distribution of the aquatic stages of mosquitoes within their breeding sites. 

I agree with the last paragraph, 

The contributions are very helpful and informative. Let me add this, you may calculate larval density based on the number of larvae per 10 dips. That statistically makes sense and is used for the surveillance of aquatic stages in medical entomology, a good reference is "Mosquito Ecology field sampling method".

I agree with these advices; however the accuracy of the counting method depends of the purpose of the study. We used to perform abacus samplings for density estimations. This method saves time because it is based on visual comparison. It is explained in this paper: "Carron, A., Duchet, C., Gaven, B., Lagneau, C., 2003. An easy field method for estimating the abundance of culicid larval instars. J. Am. Mosq. Control Assoc. 19, 353–360."

Thanks guys. I have learnt A lot

This is another  helpful refference for your need

Bull Entomol Res. 2008 Aug;98(4):323-32. doi: 10.1017/S0007485308005701.

A standardized sampling method to estimate mosquito richness and abundance for research and public health surveillance programmes.

Mendoza F1, Ibáñez-Bernal S, Cabrero-Sañudo FJ.

Just be very strict with the concept of "random dip" collection because the distribution of larvae is never homogeneous in a breeding site. You can pick 80% of the larvae with your first dip and wrongly extrapolate after.

I agree with these contributions  but  just to emphasise that during scooping, ensure that the shadows of the collectors does not interfere with the breeding site as the larvea will move away to hide giving a wrong estimate per dip.Let the shadow be away from the breeding site  and also the vibrations on the ground made by the collectors as they advance to the habitat may also affect the random sample when the larvea respond by hiding.

Thank you Sumodan, thank you Kirunda for the contributions.

The mosquito larvae estimation depend on the species you would like to count, for Aedes for example there was more than one density estimation method and indicators, such as stegomaya indices and Breautue indices. For Anopheles as well you can use the dipping estimates of 5-30 dipps depend on the size of the breeding sites.

Thanks a million for these informative responses, especially the references; Dr Kweka a zillion thanks for the attachment, I would not have had access to the article.

Just to clarify that estimating immatures' population in experimental and man made breeding sites (relatively small in size) differs when dealing with natural breeding sites that are large & difficult to handle. A transfer disposable plastic spoon can help count all immatures in artificial or experimental breeding sites. In a natural breeding site, the collector has to go round the edges to collect from multiple spots then, affix the laddle to a long stick to collect at the center of the pool, especially if there is vegetation to collect all the larvae that float or attach to any substrate; avoid shadowing that may disturb the luminosity in the collecting axis so immatures don't move.

A good discussion. But, does the same apply to avian malaria mosquitoes? That's my interest. Does anyone have identification key to this group of mosquitoes for East Africa/Africa? Any papers welcome too. It's a new area for me but have the zeal to take it up. Thanks in advance.

Thanks for  this  good discussion.Time for scooping should not be so early  morning and not late  of the day , weather is a factor, and length of the scoop.  

The use of standard dipper (350ml) recomendaded by WHO is recommendade

Use of standerd scoop is a factor as well as size of the pool.. it is recommended that  if the size is 1sq. meter 5 scoops  can be used to estimate.