Big change between tests

[SuncrestReef]

Eli,

In your first round of testing, my tank had 614 microbial types and was the most diverse of all your samples.  On the second round, my results dropped dramatically to only 213 microbial types.  Did you see similar large swings in other samples?  I haven't really changed anything in my tank or with my maintenance routine, so I'm at a loss for why it would change so much.  Any thoughts?

May 2019:

October 2019:

@EMeyer

[Lexinverts]

It’s because of the new sampling protocol that now uses biofilm sampling. The same thing happened to my tanks. 
We can only compare round 2 to future measurements using the new protocol. We can’t compare round 1 with future measurements, just to other samples collected during round 1. 

[EMeyer]

The best answer is sampling was so different we should not compare them on the basis of raw number detected. This is one of the reasons I was so hesitant to adjust the sampling method, but felt it was worth it in the end.  

I am working on a standardized diversity score that will make it possible to compare the diversity numbers themselves.

With that said, putting aside the diversity numbers themselves, the community barplots are directly comparable. I've showed before that taking two samples side by side from a single tank produces nearly identical results, so we know the measurement itself is reproducible. Your samples had plenty of DNA and sequenced well, so I have no reason to doubt the new community profile or the old one. Lacking any other information I have to conclude the community has changed and am also curious why. 

I do note that in many has it hasnt:

• You consistently have detectable levels of AOB and NOB in both samples, something many of us envy. 
• You consistently have low levels of cyano, which in both cases includes Ulvophyceae. 
• You consistently have no fish pathogens (found in 1/8 of tanks) or coral pathogens (found in 1/10 from the first round, zero client samples from the second)
• Many of the core families are consistently similar in your tank to the typical tank, e.g. in both samples, these families match the abundance in the average tank: Rhodobacteraceae, Vibrionaceae, Pseudoalteromonadaceae, Oceanospirillaceae, Hyphomicrobiaceae, Cenarchaeaceae, and Bacteriovoracaceae. That is a lot of agreement! (In comparing these plots, I note that I did change the color scheme slightly in the updated version, sorry about that. I'm trying to minimize changes to make these comparisons less confusing)

It looks to me like what has happened is a proliferation of Alteromonadaceae and a reduction in Flavobacteriaceae and Pelagibacteraceae. Since the latter two are the most abundant in the typical reef tank (and in your first sample), and Flavobacteriaceae is one of the more diverse groups, these changes alone could account for a lot of what we're seeing in your scores. 

The high levels of Alteromonadaceae in your sample are almost entirely a single type, a bacterium that has not been identified beyond the family level. It showed up in many of the tanks surveyed, but was absent from most. A few had this same exact bug at high levels. Its present in all 4 of my home tanks at relatively low levels, for example. I can't find examples of this exact bug in previous environmental samples, although other members of the family are widespread in marine environments

https://obis.org/taxon/393035

I include this just to say its not odd to have it present, but interesting to have it bloom like this in a few tanks.

Your water conditions appear to have very stable during this time, in terms of what we typically measure. I wonder if there were other unmeasured changes. By any chance have you done any ICP during this period?

Edited December 1, 2019 by EMeyer

[EMeyer]

4 minutes ago, Lexinverts said:

It’s because of the new sampling protocol that now uses biofilm sampling. The same thing happened to my tanks. 
We can only compare round 2 to future measurements using the new protocol. We can’t compare round 1 with future measurements, just to other samples collected during round 1. 

Yeah, thats definitely the safest answer. 

I do plan to make a standardized diversity score (based on resampling to a standardized sequencing depth), so we can (within the PNWMAS group) compare our round 1 and 2 samples diversity using that metric. Directly comparing the reports could be misleading. Although as I described above in many ways these results are consistent.