COUNTING LARVAE FOR DENSITY
From: Mike Corser
To: Shrimp@onelist.com
Sent: March 31, 2000
QUESTION:
I have heard of and used various methods of counting larvae to determine the actual stocking density of a pond or determining the quantity of larvae received from a laboratory.
-1st method: Release the larvae into a 200 liter tank and swirl the water with your two arms until you think enough agitation is achieved to uniformly distribute the larvae; then take four 50 or 100 ml samples of the water/larvae. Do counts, hope the distribution was uniform, take the average and extrapolate to determine the number of larvae in the 200 liter tank.
-2nd method: Do the same as above, except use excessive aeration to get a boiling effect in the tank in order to evenly distribute the larvae; then take the samples.
-3rd method: Harvest the larvae into a net and weigh the larvae; then take a 1 gram sample, count the larvae and extrapolate.
Of all these methods perhaps the third is more accurate, but overly stresses the larvae. The first two are not very accurate and depend too much on how well the larvae is distributed and how representative the sample cup is filled.
Does anyone know of more accurate ways to sample with less stress, less time and more accuracy?
Mike Corser
Pesca, S.A., Guatemala
mcorser@infovia.com.gt
********************
COMMENTS 1:
Counting methodology is controversial, and proponents of each method
will generally defend theirs as the best. Here's my opinion.
I would group the methodologies into 3 slightly different categories:
1. Volumetric sub-sampling.
2. Reduction.
3. Gravimetric (weighing).
The reduction method, which also uses volumetric sub-sampling, is my
least favorite, and is only suitable (in my opinion) for counting larvae
that will be packed into small transport containers (e.g., ice chests).
I am also not too fond of the gravimetric method, although it is very
popular in many countries (e.g., Ecuador). It is stressful to the
larvae
(especially if they are molting), and you have to correct for
moisture content, which also is not easy to standardize. I would use the
gravimetric method for very large PL's, or if you cannot chill the water.
My preference is the volumetric sub-sampling method, which is
identical to your "methods No. 1 & 2". There are a number of
techniques that you can use to increase the validity of the volumetric
sub-sampling method:
1. Chill the water before counting. This is very important,
especially
for large, active PL's.
2. Use as large a sampling beaker as possible (100-250 ml) to
increase
the sample size.
3. Count in as small a tank as is logistically feasible (200-500
L) in
order to make it easier to mix (homogenize) the larvae.
4. Count/sample in round, circular tanks; not square tanks
(corners interfere with homogenization).
5. Count/sample at the highest density possible. This reduces
sampling variability.
6. Do not stop mixing as you collect the sub-samples.
7. Collect a minimum of three samples (ideally five) per
counting tank
from different locations and depths in the tank.
I have found that turbulent aeration is not necessary, and may be
counterproductive and stressful to the PL's. The technique that I prefer
is to have two persons mixing simultaneously, each one holding small,
inexpensive, plastic (round/soft edge) dinner plates that act as paddles
(move them forcefully but not churning too hard). Avoid hand movements
that will create centrifugal vortexes in the center of the counting tank
(which will cause the PL's to concentrate in the vortex).
I have experimented with piston sampling tubes that sample a fixed
volume, and seal to prevent escape of the PL's, but I never could find one
that did not mutilate a small but significant percentage of the PL's.
There is a very simple method for evaluating your mixing and sampling
technique. Calculate the coefficient of variation (standard deviation
divided by the mean) for the sub-samples. No matter what technique you are
using, if your CV is < 5%, you are doing a great job; 5-10% is quite
good; 11-15% is acceptable; 15% is not acceptable.
Although I have worked in hatcheries, I am basically a pond guy. I am
sure that the career hatchery people on the list have strong opinions
about their techniques.
Henry C. Clifford
Technical Director
Super Shrimp Group
hcclifford@aol.com
********************
COMMENTS 2:
In Thailand, I have seen a thimble used. The thimble/small cup has holes drilled into it, so it is a sieve, and holds enough PL's for one bag. The user fills the thimble with PL's from a net, while the bulk of the PL's are held in an aerated bin. From about 100 bags, about 5 are checked - every PL is counted - and from this the total number of PL's is counted.
Here in South Africa we have an adaptation of method 1, but have never checked the accuracy of our method compared to other methods:
SA Method: Release the larvae into a 100 liter tank (up to 600,000) and swirl the water with a dish scoop until you think enough agitation is achieved to uniformly distribute the larvae; then take 2 samples of 250 (270) ml of the water/larvae. Each goes into a 10 l bucket. Using the scoop, stir the bucket water vertically vigorously and using the 270 ml wide neck bottle, take 5 samples from each bucket. If the counts from the two buckets differ widely, take another sample from the main bin to corroborate which is more accurate. Take the average and extrapolate to determine the number of larvae in the 100 liter tank. Avg. of 10 counts / 270 X 10,000 / 270 X 100,000 = PL numbers. We have used this method for 10 years.
Laurence Evans
laurence@amatikulu.co.za
***********************
COMMENTS 3 :
For a discussion on volumetric methods for counting PLs at the hatchery
you may want to check:
Juarez, L.M., A.H. Luxem, and D.B. Rouse. 1996. Sampling shrimp
populations in Hatcheries. Journal of the World Aquaculture Society, Vol.
27, No. 2. pp. 218-222.
Lorenzo Juarez
ljuarez@seafarmsgroup.com