As the summer winds down, it looks as though the project worked better than I had hoped.  There is a lot left to do, so this is far from the end, but what a great beginning!

To remind you of the the primary goal of the summer’s project, we wanted to use the DNA contained in the slugs’ kleptoplasts to identify their primary food plant(s).  The previous posts described how we worked out methods, collected slugs and candidate food algae, extracted the DNA, amplified the rbcL gene from the chloroplasts, and sent it off for sequencing.

The first sequence that came back from Macrogen did not look very good, which was disheartening.  The chromatograms looked awful, and the sequence was gibberish, causing concern that our extractions or PCR reactions were contaminated.

Lousy chromatogram from Sanger sequencing. Note multiple possible bases (different colors) at each site. Uninterpretable.

Nonetheless, Paul Kim at Macrogen promised to optimize the reaction and sequencing conditions, and worked hard to provide interpretable data. Patience and persistence have finally paid off, and we can make some simple, declarative statements about the slugs and their food plants.

Codium sequence from Macrogen. Note a few sites showing more than one possible base, presumably polymorphisms.  8/10/16

Statement 1: We obtained usable rbcL DNA sequence from Codium, Ulva and Elysia.

Statement 2: Elysia diomedea steals most, if not all of its kleptoplasts from Codium.

To flesh out these statements a bit:

From Bahia, we now have DNA sequence for Codium simulans and for Ulva.  The Codium data is the first for the species.  Although rbcL sequence for related species (such as C. isabelae) can be found in the NCBI database, there is currently nothing for C. simulans.  We’re not sure which species of Ulva we used, although it is likely to be Ulva californica.  In theory the DNA sequence could have told us which species it was, but the region of the rbcL gene that we amplified and sequenced is identical to that in many of the species in the database, so we would need to try another gene, or a different region of rbcL.  An important lesson from this year’s work was that we need to preserve samples of the algae we sequenced.

The most exciting result was that we got sequence from E. diomedea kleptoplasts!  Overall, we extracted DNA from two individual slugs at different times, and performed at least three separate PCR amplifications (both in BLA and at USG when I got back), and they all came back matching Codium!  In retrospect, it is not a shock that slugs that we found in close association with Codium, and which spend a lot of their free time on Codium, actually eat Codium.  

Portion of rbcL sequences extracted from Codium simulans (top), Elysia diomedea (middle) and Ulva sp. (bottom). Sites at which Ulva differs from both Codium and Elysia are indicated by arrows.

The figure above shows a small portion of the sequence, highlighting a few of the sites at which Elysia and Codium differ from Ulva.  Overall, the DNA sequence from Elysia was 99% identical with that of Codium, and those few sites that differed appeared to be locations at which there was variation between individuals.  Ulva showed about 81% identity to Codium and to kleptoplasts from Elysia.

E. Diomedea on Codium in tank at BLA station

Despite how it sounds, this is not a trivial result.

First off, Codium has been suspected, but never confirmed as a the food plant. Back in 1969, Trench and colleagues said that E. diomedea fed on green algae, possibly C. simulans, based on the chlorophylls found in the slugs and the morphology of the kleptoplasts, but their methods could not reliably distinguish between green algae species.

As a corollary, there is no evidence that they eat Ulva or Padina, despite being surrounded by them.  We did not get rbcL sequence from Padina this year, but it is not closely related to Codium, and the sequence in the database for P. durvillei (the most common species in our study area) shows roughly 70% identity to that from Codium and E. diomedea.  Had there been significant Padina or Ulva DNA in the slug sample, the presence of multiple divergent sequences are likely to have made interpreting the results impossible.  In other words, we got lucky that there was one dominant species of kleptoplasts.   Having sampled only two slugs, we can’t rule out other food plants.  Another caveat is that the result shows that chloroplasts from Codium persist in the slugs’ tissues, but the slugs could be eating other species for which the chloroplasts do not last as long inside the slugs.

Another important conclusion is that our methods actually worked.  As a neurophysiologist setting up a molecular lab in a dusty, hot garage in an isolated location, there were no guarantees that we would get any usable data.  In addition, we used degenerate primers for PCR, to amplify rbcL sequences from all potential algae species, counting on DNA sequencing to tell us which species were present.  Our choice of Sanger sequencing, which is much less expensive but prone to problems if the amplified DNA comes from more than one species, could have also caused complications.  Planning, persistence, and some luck all worked in our favor.

With these data in hand, there is lots more to do.  To fill in some of the gaps discussed above, we need to sample from more slugs in more locations.  At the same time, we need to more systematically collect specimens and DNA from algae at different sites around the bay, especially C. simulans.  If we are going to generate DNA sequences, we may as well do it in such a way that we can add them to the database.

There is also a lot to be done to understand the big picture of kleptoplasty and how E. diomedea fits into the ecology of the bay.  Because of delays in receiving equipment, we had very little time to prepare the behavioral experiments before we left Maryland.  On top of that, the losses and stress caused to the slugs by the extreme heat this year, resulted in essentially no data regarding the slugs’ preferences for light.  The I-mazes are build and ready, and we plan to add a chiller to the holding system, so procedures should be perfected before the next field season.  We also still don’t know much about their environmental requirements.  They eat Codium, and live on Codium, but do they have other requirements in terms of water movement, temperature, nutrients, or turbidity?

That the project worked can be chalked up to a lot of planning, hard work, and generosity on the part of a great group of people.  At the risk of sounding like an Academy Award acceptance speech…

Photobiology Group. Cristal, Rosalia, Nancy, Richy, Allison, Susan & Thiago.

There would have been no Photobiology group without the “Angels,” Cristal, Rosalia, Nancy, Allison, and Susan.  It was so much fun to watch them work and learn.  They will be giving their presentation during the Report to the Community for Ocean Discovery this week, and it will be great.

Richy Alvarez, the intelligent and talented Directed Research Fellow, was another reason this project came together.  There are so many big and small things that he did to make sure equipment was ready and that the students were prepared, I can’t thank him enough.  Big thanks also to Thiago Lima, for generously taking time away from his postdoc at Scripps to work with the students in the field, and for giving advice on the project (he is an actual molecular biologist) along the way.

Huge thanks to all of the staff at Ocean Discovery Institute, especially Joel Barkan, who coordinated the process of turning the plan into a reality when I was 3,000 miles away.  I can’t say enough good things about the support I received from everyone at Ocean Discovery, at all levels, and how easy it was to work so closely with so many people.  Bahia de los Angeles is a magical place, but doing science there can be a hot, tiring affair.  Working with this group makes the process so much more fun.

There would have been no time to work out procedures once we arrived in Bahia, so Maryam and Haseeb’s experiments and troubleshooting at USG were crucial.

The experiments also required equipment.  Some, like the PCR machine and centrifuge, were generously loaned (thanks ThermoFisher and USD!).  Others, such as the tanks and DNA sequencing were purchased from vendors who went the extra mile to do things well and on time (Glasscages and Macrogen).

None of this could have happened without permission from the Comisión Nacional de Áreas Naturales Protegidas (CONANP), which administers the Biosphere Reserve at BLA, and the support of Jose Mercado, who owns and operates the Casa Caguama field station in BLA.

Dr Drew Talley, BLA staff office 7/1/16

Finally, I owe an enormous debt to Drew Talley, my best friend for over 40 years.  He introduced me to Bahia many years ago, and worked tirelessly this year to secure loans of equipment, permits, and who has been incredibly supportive of the development of this project.  He has the right to call himself the Captain.