Dear Maria,

1. and 3. Do you have at least a piece of mesh with a mesh size of 10 or 20 µm? If so, you can easily build a device to filter water and concentrate your sample. Take a water jerrycan and cut off the base. Then make a hole in the cap, keeping only the threaded ring. Place the mesh over the mouth of the jerrycan and screw the ring back on, trapping the mesh. Now you pour water through the open base, and the water that comes out through the mouth is filtered by the mesh. In this way, you obtain a concentrate on the mesh. Remove the mesh once the jerrycan is empty, and you'll have your concentrated sample.

Another option is reverse filtration, which is useful if you are working with very delicate phytoplankton, since it avoids applying pressure on algae. For this, pour the water to be concentrated into a large beaker. Separately, prepare the same device described above, but using a small bottle instead of a jerrycan. Submerge the bottle, with the mesh end facing downward, into the beaker water. Water will enter the device through the mesh at the bottom, due to the principle of communicating vessels. Once the bottle is full of filtered water, remove it quickly from the beaker to minimize water loss, and discard that water. Repeat the process many times. As a result, the phytoplankton in the beaker will become increasingly concentrated.

As you sugegst, you may let the sample settle, as well. You would get a concentrated sample this way. But it takes too long, does not allow to concentrate much water and does not permit to get rid of small debris and sediment.

Anyway, if you need to count phytoplankton, you should not use such concentration strategies (because you lose small cells and you would not be certain about how much water you are analysing actually). I would count phytoplancton on raw water (not concentrated) instead.

2. This depends on the purpose of your work, the size of the river, and the specific section you're studying. If it's a lower reach, you might be dealing with transitional water, where seawater can intrude. In such cases, you should consider the possibility of a layer of saline water beneath the freshwater. If it's an upper reach, it might be more useful to study the benthic community (especially diatoms) rather than plankton, since benthic organisms tell you what is happening in the site itself, while plankton reflects what has happened upstream. And the fact that you mention sampling upstream and downstream suggests you're trying to assess the impact of some disturbance the river is experiencing. This is not an easy question to answer without more context...

Thank you, Mr. Gonzalo for the suggestions. This will be very helpful in fine-tuning our methods. I was doubting that there would be a high chance of error if we were to just let the sample settle. I am now considering the reverse filtration method, but would a 25 micron mesh suffice? I am currently scouring online for available nylon mesh filters accessible in our area and this is the smallest size I can find.

For more context about our study, we plan to assess the health of the river through determining the specific phytoplanktons in parts along the river (upstream, Mid, downstream). The river is approximately 54 km long.

1. Indeed for the downstream, it is close to sea so saltwater may invade the area.

2. Due to limitations in equipment, we were previously considering to collect samples on the upper layer of water. Our lab only has a planktonet with mesh sizes suitable for larger zooplanktons only, so we may not be able to use it as is.

3. Do you have any suggestions on how deep we should take samples and how? I fear if we simply submerged the bottles to fill with water, it may not contain accurate sample/representation of the site itself.

For phytoplankton, I first preserve a sample in glutaraldehyde (0.25 to 0.5%), approximately 100 mL or 1000 mL, depending on the concentration of organisms of raw water. I then use a graduated cylinder and measure an exact amount - the 100 mL or 1000 mL, then add enough Lugol's solution to turn the sample tea colored. Lugol's binds to the cells and allows them to settle. Air bubbles might form - and I lift the cylinder and bang it on a hard surface, often a few times, to get out these air bubbles. Allow 1 hour of settling time per cm of vertical length of in the graduate cylinder. I typically give the sample 24 hr. The next step is to remove the upper 90% of the liquid by aspiration. I use a vacuum pump and a hose ending in a pipette, with a jar collecting the liquid. This leaves you with a 10-x concentration of algae in the bottom of the graduated cylinder. For example, 100 mL down to 10 mL. If you settle 1000 mL down to 100, you may also do a second settling of the 100 down to 10 mL, which would give you a 100x concentration. It is up to you to determine if the dilution from the preservation with Lugol's needs to be accounted for.

The concentrated sample is vigorously shaken in the graduated cylinder and poured into smaller bottle or centrifuge tube, which is labeled and can be tightly closed. When I use a centrifuge tube, I get a final value of the volume which is used in a simple equation to calculate the concentration factor.

Some organisms, like the Cyanobacteria with gas vesicles, like Microcystis, are difficult to settle, even after treatment with Lugol's. If this happens, I recommend capturing the floating colonies using a pipette and adding them to the concentrate. Always record the final volume of the concentrate.