two species of marine diatoms

Franz Neidl · Post by **Franz Neidl** » Fri Jan 07, 2011 7:40 am

I dont know the species of this marine diatoms. The genus of the second diatom is Bacteriastrum.
Somebody could ask for what are the bristles (spines) good? This is the answer:

"All plankton must avoid sinking. Phytoplankton need sunlight for photosynthesis, so they must stay within the photic zone, usually the top 100 meters (MARE, 1995). Zooplankton depend on phytoplankton and other zooplankton for food, so they must avoid sinking as well. Plankton avoid sinking by increasing their surface area and/or decreasing their density. Most plankton are quite small and so have a larger surface area to volume ratio than do larger organisms. Flattened bodies and appendages, spines, and other body projections also slow sinking by adding surface area without increasing density (MARE, 1995). Some diatoms resist sinking by forming chains". (from: http://marinediscovery.arizona.edu/less ... ans/2.html )

Objective 20x Ph, stacks

Franz

BJ · Post by BJ » Sat Jan 08, 2011 5:59 am

Franz,

I particularly like your Chaetoceros photo ; I am fairly certain that it is C. curvisetus:

http://www.algaebase.org/search/species ... s_id=37414

Perhaps René will confirm (or not ! )

The only species of Bacteriastrum with so many spines per cell is B. hyalinum

The photo of Bacteriastrum has a shadow along both sides of the chain...do you think this is an artifact from the stacking process..perhaps a problem for the software trying to make sense of so many spines ?

You might enjoy this article which discusses why diatoms have silica walls....they come to perhaps rather surprising conclusions:

http://onlinelibrary.wiley.com/doi/10.1 ... 1022.x/pdf

kind regards,

Brian

Wim van Egmond · Post by **Wim van Egmond** » Sat Jan 08, 2011 9:02 am

Nice work Franz and an interesting comment about the reason why they form these shapes. I wonder if there is also the extra advantage of size which prevents them from being eaten. I can imagine it has several benefits. But when I see Chaetoceros in my samples it sinks slower than most other diatoms.

Wim

René · Post by **René** » Sun Jan 09, 2011 3:30 am

Yes, Wim is right, besides of lower sinking rate, avoidance of easily being preyed upon is a strong selection criterium during evolution. It is not only limited to predators perse, it is also known that Chaetoceros blooms can kill fish because of irritation of the gills by the spines.

Difficult to assign this Chaetocers, it could be one of 3. Without a closeup of the aperture between cells, I can only say that the most likely choice is C debilis, based on the (smallish) diameter of the curves.
Best wishes, Rene

BJ · Post by BJ » Sun Jan 09, 2011 5:05 am

Hi,

René wrote:

"......avoidance of easily being preyed upon is a strong selection criterium during evolution. It is not only limited to predators perse, it is also known that Chaetoceros blooms can kill fish because of irritation of the gills by the spines. "

but this is probably not advantageous to the Chaetoceros as many fish would be predators of the herbivores (say copepods) which would eat the alga. Therefore less fish = more herbivores = more grazing.

Nature is so very complex!

regards,
Brian

Franz Neidl · Post by **Franz Neidl** » Mon Jan 10, 2011 11:15 am

Thank you Brian, Renè and Wim for your interesting discussion and for your help with identification.
for Brian: yes, this (the shadow along both sides of the chain) is an artefact coming from the stacking process. I have this generally when I have a strong halo-effect in the object. Next time I will ask Rik what to do in this case...

Franz

rjlittlefield · Post by **rjlittlefield** » Mon Jan 10, 2011 12:15 pm

Franz Neidl wrote:this (the shadow along both sides of the chain) is an artefact coming from the stacking process. I have this generally when I have a strong halo-effect in the object. Next time I will ask Rik what to do in this case...

This sort of halo with this sort of subject is very difficult to get rid of. I assume you are using PMax, in which case the halo is essentially a weak and often inverted copy of information from out-of-focus frames. The DMap method does not give this sort of halo, but it is often difficult to get DMap tuned to work well for high magnification microscope work.

You might try processing in DMap using radius settings that are much larger than default, something in the vicinity of 1/200 of the image width in pixels. If the images shown here are processed at 4000 pixels, then an Estimation Radius of 20 and a Smoothing Radius of 10 might work well. But it is impossible to predict this accurately, so you might need to try a range of values.

Of course there is also the trick of using retouching to merge intermediate outputs. For example it may be easy to tune DMap to handle the body but not the spines, and vice versa, or perhaps the body with DMap but the spines by PMax. If so, then some judicious brushing to combine the two DMap outputs, or a DMap and a PMax, might give a clean result without too much work.

--Rik