"Biodiversity is dead, long live biodiversity" 10 month microbiome data from BRStv.

[taricha]

Update on this.... Q of why aquabiomics showed that new material could reliably increase microbial biodiversity, but BRS results showed that none of their systems acheived average diversity very quickly...

Idea 5b: The other big headscratcher (for me) in the BRS results.
Aquabiomics demonstrated in two separate articles (with replicate tanks) that it was possible to add material either to a new system...
Establishing a Healthy Microbiome in a New Aquarium Using Live Rock

...or to supplement an established system...
Effects of live sand & mud on the microbial communities in my tanks
...to quickly (in <4 weeks) increase the diversity of the system to >50 percentile.

But BRS results were that zero out of 12 tanks acheived results like that in 4 weeks - only 1 tank of 12 got to 50th percentile by week 15.

I wanted to see if there were examples of other people replicating aquabiomics results, and I found two that actually did before/after with added live material.
example 1
before : number = 300 , percentile = 36
after Live sand from IPSF: number = 488 , percentile = 86

example 2
before: number = 162, percentile = 05
after live sand also from IPSF: number = 726, percentile = 99

In other words, it looks like the aquabiomics results and not at all like the BRS ones.

so the answer to this question is probably yes, aquabiomics results were repeatable....

As you can see, All the BRS tanks moved similarly on this scale, and none of the BRS tanks got anywhere near the measured diversity that aquabiomics found repeatedly from adding the material in its tests.

So a couple of different possible interpretations:
Does this mean that the Aquabiomics results were not as replicable as they appeared? Have other people been able to raise their measured microbial diversity dramatically with those materials? (maybe @AquaBiomics could comment)

It looks like this is more likely correct interpretation - not all live sand / mud / rock is remotely the same in ability to create measurable diversity, and no actual high diversity material made it into the BRS tests.

I think it more likely that the BRS starting media were actually significantly less bacterially diverse than the material that aquabiomics used in their testing.
This is a simpler explanation to me: that the BRS live mud (aquaforest) was not similar to the live mud/sand from floridapets that aquabiomics used, and that the live rock that BRS got was more like the aquabiomics live rock A, than their high diversity live rock B.

 

[ReefGeezer]

I wonder if TBS or a similar outfit would have a large enough interest in these kinds of experiments to either aid in running them or to donate material for the tests.

If the tests turn out beneficial to them, it’s good advertising. If they don’t… I dunno

I'd imagine there are a lot of variables that would determine the outcome of the testing. Some would be hard to control, others hard to identify. False assumptions could easily be made. I would not want to risk a negative outcome.

I think this thread has shown that live rock might just be cheap wine in a fancy bottle with a high price tag. method.

I think it simply demonstrates the limits of the method AND once again, puts a premium on microbial diversity/balance rather than concentrating on overall diversity of all the organisms in the system. Do you know why that happens? I think I do... Retailers can make money off of products and services associated with microbes but not the other organisms.

 

[Dan_P]

I tricked @Dan_P into thinking it was interesting enough that he took the image of the bar charts from the video and converted the major families into numerical data.

@taricha has what I call the Tom Sawyer Effect. This not the first time I paid him to whitewash his fence

 

[taricha]

Analogous to your “uglies” score, have you considered a score for the inoculum size? The control might be 0 while the Gulf rock might be a 10, along with the 360 rock and sand. Hard to judge the Indo rock from its appearance. I ask because I have been wondering if community development is inoculum size dependent.

I can for some. Not sure if I can get a sensible number for all. I can at least do 3 groups - Hi, low, and can't determine.

 

[taricha]

Here's my shot at trying to put some numbers to the idea of stability in these BRS tanks.
This is an example of a measurable thing that (with few exceptions - big stable population of pathogens/bad guys) most would agree represents something important. If bacteria matter to reef tanks at all, and if reef tanks do better with fewer dramatic changes - then it's logical to think you'd prefer the major bacterial families stay more stable, rather than regularly spiking and zero-ing.

Below charts show the CHANGE in each major family prevalence between the tests times: from week 2 to 4 (blue), week 4 to 10 (orange), and week 10 to 15 (gray).
The final set in each chart is the AVERAGE change across the major families. I split it into two charts, the higher change tanks in one chart and the lower ones in the second.
"Higher change" group

The highest changing systems (weeks 4->10->15) were the Biobrick, and 2 cups of sand.

"Lower change" group

The lowest changing systems (weeks 4->10->15) were Gulf Rock and Indo Rock.

Spoiler - none are very stable. If there are 5 or 6 major family groups and a total of 1.0 then the average family size is around 0.15-0.20 - and the average family change in each time period across most tanks was around ~0.10. So roughly speaking, on average, families in all these tanks are changing by about half or more of their prevalence. The most "stable" tanks weren't much lower than that.

I think most people would care more about the stability from weeks 4 to 10 to 15. rather than weeks 2 to 4, when dramatic changes might be more expected. And most tanks did have larger changes from weeks 2 to 4. But it's not the case that overall most tanks were more stable from week 10 to 15 than they were from week 4 to 10. Some were, others not.

 

[Dan_P]

Here's my shot at trying to put some numbers to the idea of stability in these BRS tanks.
This is an example of a measurable thing that (with few exceptions - big stable population of pathogens/bad guys) most would agree represents something important. If bacteria matter to reef tanks at all, and if reef tanks do better with fewer dramatic changes - then it's logical to think you'd prefer the major bacterial families stay more stable, rather than regularly spiking and zero-ing.

Below charts show the CHANGE in each major family prevalence between the tests times: from week 2 to 4 (blue), week 4 to 10 (orange), and week 10 to 15 (gray).
The final set in each chart is the AVERAGE change across the major families. I split it into two charts, the higher change tanks in one chart and the lower ones in the second.

The highest changing systems (weeks 4->10->15) were the Biobrick, and 2 cups of sand.

The lowest changing systems (weeks 4->10->15) were Gulf Rock and Indo Rock.

Spoiler - none are very stable. If there are 5 or 6 major family groups and a total of 1.0 then the average family size is around 0.15-0.20 - and the average family change in each time period across most tanks was around ~0.10. So roughly speaking, on average, families in all these tanks are changing by about half or more of their prevalence. The most "stable" tanks weren't much lower than that.

I think most people would care more about the stability from weeks 4 to 10 to 15. rather than weeks 2 to 4, when dramatic changes might be more expected. And most tanks did have larger changes from weeks 2 to 4. But it's not the case that overall most tanks were more stable from week 10 to 15 than they were from week 4 to 10. Some were, others not.

Two (at least) factors were not considered and therefore not measured: N,P concentrations (maybe even the amounts of inorganic vs organic nitrogen) and the algae biome. I think this was another big fail for the BRS experiment, making data interpretation similar to that of an airline crash investigation. Because the number of species network interactions is so huge, we are unlikely to squeeze out much more information than you already have. Time to plan the next experiment maybe, at least our dream experiment.

 

[sixty_reefer]

Two (at least) factors were not considered and therefore not measured: N,P concentrations (maybe even the amounts of inorganic vs organic nitrogen) and the algae biome. I think this was another big fail for the BRS experiment, making data interpretation similar to that of an airline crash investigation. Because the number of species network interactions is so huge, we are unlikely to squeeze out much more information than you already have. Time to plan the next experiment maybe, at least our dream experiment.

I agree that the CNP residual would have gave a better picture, ideally doing this sort of testing in conjunction with a N-Doc would have been much more beneficial to try and have some sense of the influence of nutrients on bacteria since that’s their main source of energy.
It’s a known factor that phosphates is absorbed and released by many aragonite surfaces more in particular during the first few months wile trying to balance itself with the surfaces, in some systems the phosphates could have been more available and in other depleted.

 

[taricha]

Two (at least) factors were not considered and therefore not measured: N,P concentrations (maybe even the amounts of inorganic vs organic nitrogen) and the algae biome.

While I partially agree, let me take the opposite side for the sake of discussion, and thinking through any future tests along these lines.

Nutrient measurements here aren't really low-hanging fruit. NO3 & PO4 aren't actually the driving nutrients for these bacterial families. If you go through the necessary compounds to feed these bacterial families (aquabiomics family info) it's stuff like one group likes carbohydrates, another needs methylated amines, another likes organic acids, another does better on blank seawater controls than any nutrient addition at all. So unless you are doing N-DOC or your own DIY version we aren't going to catch the bacterial foods anyway. Also you and I generally agree on the basic proposition that the driving conditions for growth are more often varied and local on surfaces rather than uniform in the bulk water, so even a test of organic N and C in water likely isn't catching what's driving most of these communities. Also, an organic test for C and N isn't telling you the form of it (which is theoretically what drives the families) - only the totals. A much better marker for is what you said - the algae community, because the photosynthetic community (aside from the food input) is the source of many of these compounds. And short of something really clever - pics and video of what grows in the tank (which we have) is the best marker for these driving effects.

--

1 - Indo live rock

2 - Dry rock/sand

3 - Live (bag) sand

4 - tank sand

5-coral

6-dark rubble

7-tank rock/sand

8-TankWater

9-gulf wet rock

10-artificial rock saltwater cured

11-reef mud

12-biobrick

Uglies score

8

1

0

6

4

3

4

5

9

1

2

6

--

chryso

hair

hair

flatworms

cyano

hair

hair

cyano

cyano

brown dust

GHA

--

hair

diatoms

cyano

hair

hair

diatom

--

dino

we are unlikely to squeeze out much more information than you already have. Time to plan the next experiment maybe, at least our dream experiment.

I'm almost there - just a few more things I want to look at before I'll say I've exhausted my data mining efforts here:
Water vs Surface families
High nutrient vs low Nutrient families
Individual family groups

I'm curious to see if there's strong connections between data on any of the above and the obvious observables about these systems.

 

[areefer01]

So how hard is it to update the data and regen the graphs?

If easy, who has a recent test result that is say of a display 5 years or older. Plug that data in and how does it look.

 

[taricha]

So how hard is it to update the data and regen the graphs?

If easy, who has a recent test result that is say of a display 5 years or older. Plug that data in and how does it look.

BRS also gave results of three of their established display tanks, for comparison.
Generally, I would say the displays look more different than similar to these 12 test tanks.
Which is one of my reasons for wondering if maybe stuff takes a really long time, because none of the systems in 15 weeks looked very much like the main display systems as far as aquabiomics is concerned.
To say it another way, the 12 test tank microbial communities looked more like each other than any of them looked like any of the three established displays.

 

[areefer01]

BRS also gave results of three of their established display tanks, for comparison.
Generally, I would say the displays look more different than similar to these 12 test tanks.
Which is one of my reasons for wondering if maybe stuff takes a really long time, because none of the systems in 15 weeks looked very much like the main display systems as far as aquabiomics is concerned.
To say it another way, the 12 test tank microbial communities looked more like each other than any of them looked like any of the three established displays.

That is what I would have expected. I wasn't aware that data was there.

Thanks.

 

[Koleswrath]

To say it another way, the 12 test tank microbial communities looked more like each other than any of them looked like any of the three established displays.

This is what I find the most interesting......not what I would expect at the (almost) 4 month mark between these very different startup methods. I wonder if the addition of the slurry of crap they threw in there disrupted the whole microbial succession that would have naturally occurred.

 

[Dan_P]

While I partially agree, let me take the opposite side for the sake of discussion, and thinking through any future tests along these lines.

Nutrient measurements here aren't really low-hanging fruit. NO3 & PO4 aren't actually the driving nutrients for these bacterial families. If you go through the necessary compounds to feed these bacterial families (aquabiomics family info) it's stuff like one group likes carbohydrates, another needs methylated amines, another likes organic acids, another does better on blank seawater controls than any nutrient addition at all. So unless you are doing N-DOC or your own DIY version we aren't going to catch the bacterial foods anyway. Also you and I generally agree on the basic proposition that the driving conditions for growth are more often varied and local on surfaces rather than uniform in the bulk water, so even a test of organic N and C in water likely isn't catching what's driving most of these communities. Also, an organic test for C and N isn't telling you the form of it (which is theoretically what drives the families) - only the totals. A much better marker for is what you said - the algae community, because the photosynthetic community (aside from the food input) is the source of many of these compounds. And short of something really clever - pics and video of what grows in the tank (which we have) is the best marker for these driving effects.

--

1 - Indo live rock

2 - Dry rock/sand

3 - Live (bag) sand

4 - tank sand

5-coral

6-dark rubble

7-tank rock/sand

8-TankWater

9-gulf wet rock

10-artificial rock saltwater cured

11-reef mud

12-biobrick

Uglies score

8

1

0

6

4

3

4

5

9

1

2

6

--

chryso

hair

hair

flatworms

cyano

hair

hair

cyano

cyano

brown dust

GHA

--

hair

diatoms

cyano

hair

hair

diatom

--

dino

I'm almost there - just a few more things I want to look at before I'll say I've exhausted my data mining efforts here:
Water vs Surface families
High nutrient vs low Nutrient families
Individual family groups

I'm curious to see if there's strong connections between data on any of the above and the obvious observables about these systems.

Sorry, not letting you take the opposite side alone Also some ideas.

If anything, a closer accounting of CNP input could have helped us understand the BRS results. It certainly had little to do with bacteria. I am finding substantial differences between 0.3 and 1 ppm nitrate equivalent of protein added per day. Nitrate and phosphate measurements would probably have been 0 ppm each, but the tanks were still being fed.

I don’t have any bright ideas about estimating CNP inoutwhen you do something like rot Gulf or Indo rick in the dark for a month and then starve it by keeping it in a nutrient depleted system. Not sure why anyone would think rotting live rock in an aquarium wouldn’t end in a mess.

This brings me to a bright idea for putting bacteria into a system without the rotting. Bacteria are going to seed a system pretty quickly. If you think live rock has the right mix of bacteria, place it in the new aquarium for a day or two or a week, then remove it and ghost feed the system at less than 1 ppm nitrate equivalent of protein (fish flakes). This could also be done in a bucket to save water.

If aquarium water has the magic mix, fill the new system with 100% or 50% tank water for a day or two and then do a 100% water change. Bacteria are seeded but the organic junk in the water is not. Feed at the above rate.

The uglies are the price of impatience maybe. If you start a new aquarium and grow bacteria and micro algae for a month or two before adding any livestock, maybe new aquaria would start out better. GHA and other macro algae will still plague new systems until we learn to keep health grazers (that’s right, feed your clean up grew please).

 

[Dan_P]

Here is my last bit of fun with BRS data. I used the difference from control data at 15 weeks from each experiment. I plotted two family on the X and Y axis and the third was used to change the width of the dot. A dot that appears as an open circle is a negative number, i.e., there is less of the family in the control. Solid dots are positive, meaning more of the family in the control. Here is an example.

4 of the 11 experiments showed a large amount of the “pelagi” family relative to the control. The rest of the experiments showed a mix of results. I have no idea if this means anything. This was the only interesting example I found of all the three-way comparisons. Here is another example use of a 3-D plot. The red circle marks the experiments that showed a low ”taricha uglies“ score. There is also a green group and a blue group. I have no idea if the reality of this pattern is like a shape that I see in a cloud, but maybe….

 

[mmoner]

Maybe not all uglies but some should be related to C:N: P ratios somewhere? I mean in all dinoflegellates threads and videos, somewhere it is said "balancing those three in the advantage of right "bottled" strains could fix" them or "antibiotics are the cure". Maybe all we need is to look for the strains which are dominant, when they appear in tanks? (not only new setups but more examples from mature tanks) I guess BRS gave us some clues, and thanks for all graphs BTW.

 

[taricha]

Maybe not all uglies but some should be related to C:N: P ratios somewhere?
...
Maybe all we need is to look for the strains which are dominant, when they appear in tanks? (not only new setups but more examples from mature tanks) I guess BRS gave us some clues, and thanks for all graphs BTW.

So far, looking through this data I've found the story is almost always less clear than the expectations for how it "ought" to work.

This is what I find the most interesting......not what I would expect at the (almost) 4 month mark between these very different startup methods. I wonder if the addition of the slurry of crap they threw in there disrupted the whole microbial succession that would have naturally occurred.

That slurry of uglies wasn't added until after 15 weeks. So it wasn't in the microbiome-analyzing phase of the experiment. So that part isn't to blame.

 

[taricha]

Bacteria are going to seed a system pretty quickly. If you think live rock has the right mix of bacteria, place it in the new aquarium for a day or two or a week, then remove it and ghost feed the system at less than 1 ppm nitrate equivalent of protein (fish flakes). This could also be done in a bucket to save water.

If aquarium water has the magic mix, fill the new system with 100% or 50% tank water for a day or two and then do a 100% water change. Bacteria are seeded but the organic junk in the water is not. Feed at the above rate.

Are you suggesting this because you find algae growth to be stronger with inoculation times longer than a day or two, or is this thinking because of the nutrients that come along with leaving the inoculating medium in place?

 

[Dan_P]

Are you suggesting this because you find algae growth to be stronger with inoculation times longer than a day or two, or is this thinking because of the nutrients that come along with leaving the inoculating medium in place?

Multiple reasons, a bit of a convoluted story.

Scientific literature ( I have read a tiny fraction of it) teaches us clean surfaces can be colonized within an hour of submersion. The pioneers might or might not be the long term residents that form the mature biofilm, which means there is competition and late arrivals. I assume that bacterial dispersion from biofilms on solid inoculum like live rock and sand is quick and would not take more than a day or two to release enough for seeding an aquarium and not enough time to rot and overfeed the developing biofilm. Aquarium water can seed a microscope slide with algae within an hour.

My algae experiments on microscope slides indicate that the algae colonizers also arrive quickly, but 24 hours ensures enough settle to develop a mixed culture quickly (I am impatient). Longer than a day, I find that too many single cell algae “spores” have settled and start to grow rhizomes and filaments, fouling the slide and making it difficult to observe and count the single cell organisms. The 24 hours is my best guess for my aquarium water inoculum to seed the slides

When I scaled up the slide experiments to 0.2 L aquaria, the systems were inoculated for only 24 hours with aquarium water. This provided systems that completely consumed the protein in fish flakes and leaving no detectable ammonia or nitrate. I conclude heterotrophic bacteria colonized the system within the 24 hour inoculation period. Algae also began to grow within a few days indicating nitrogen was available from fish flake digestion. The algae growth was weak and accumulated at the higher dose of fish flake. The lower dose aquaria were “pretty”.

I have yet to perform the obvious experiment of taking live rock from the aquarium or rubble from the dark sump and doing a 24 hour seeding experiment. If data becomes available that certain bacteria families influence algae growth, I will also have to collect eDNA to see what the bacteria are doing. By the way, there is an open invitation for someone to try the quick dip seeding technique and discover the optimum dipping time.

 

[C4ctus99]

By the way, there is an open invitation for someone to try the quick dip seeding technique and discover the optimum dipping time.

What is the quick dipping seeding technique and how would one do it?

 

[Dan_P]

What is the quick dipping seeding technique and how would one do it?

The quick dip seeding is still an idea.

Using aquarium water to seed a new system, the technique would be to fill the system with aquarium water for twenty four hours and then do a 100% water change.

For seeding with sand (sand in very shallow dish to enable a simple removal) or rock, the big unknown for me is how long would it stay in the new system. I have not run the experiment yet. When I do, I will try several hours, 24 hours and one week. I will then look for the new system to cimpletely consume fish flakes, not accumulate nitrates and to have minimal algae growth. I might run the experiments a month or two. This means the quick dip might take a week. I guess that the sand or rock would be returned from where they came or thrown into the garden.

I have not run across any information that could help estimate how fast bacteria leave the biofilm (disperse) or under what conditions. Maybe just plunging a live rock into nutrient depleted, sterile saltwater would be enough the the bacteria to abandon the biofilm and go looking for nutrients.

Make any sense?