Intense UK-France work trip (October-November 2018)

The Seabird Sound project is already in it’s second year and I’m still happy to be in this multidisciplinary team! (Have a look here if you’re getting lost with the many names in this text!). My role here is to develop and apply methods for movement analysis to contribute to a better understanding of seabird spatial behavior and how it is shaped by infrasound and meteorological conditions.

The seabird-ecology counterpart in the project is in Liverpool (Sam and Tommy) so we agreed that I would travel there to work with them. Since I had to make such a long trip from Fort Lauderdale, I decided to make the most of it, adding several destinations and working with as many people as possible.

This is a narrative of that month-and-a-half trip to Europe, by destination:

Chizé

The Centre d’Études Biologiques de Chizé (CEBC) is a CNRS lab is known for its research in marine mammals and seabirds, and for its scientific campaigns for data collection in faraway places like Kerguelen and Crozet islands. Sam, Tommy and I met in Chizé to discuss Tommy’s ideas for his paper on wind’s role on albatross behavior during foraging trips with Henri Weimerskirch. Henri has collected part of the data being analyzed and has worked with albatrosses for many years, so his feedback on Tommy’s findings was highly valuable. We also attended the outstanding PhD defense (viva) of Julien Collet, as Sam was a member of the jury, investigating strategies seabirds use to search and reach feeding areas. During this short but intense visit (2 days), I managed to squeeze in some time to finish a manuscript reviewing R packages and submitted it literally at the last minute: Not only was it the last day to submit the paper; we were also going to miss our train back to Paris if I didn’t finish on time. I made the quickest paper submission in my life so far (about 20-30 minutes) and I hope it will remain the quickest one.

Liverpool

This was the main destination of my trip. I was going to stay there for a bit more than two weeks so I had several tasks programmed for the journey.

The other postdocs of the project, Jeff and Ollie, joined Tommy and me at the University of Liverpool for the first week. We held meetings for just the four of us, to present our different topics of research to each other: introductions to our subjects, preliminary results and perspectives. I am very much familiar with Tommy’s research, as I play an active role in it. It was amazing to be able to learn from Ollie’s and Jeff’s own research.

ollie-pressure.jpeg
High and low pressure systems

At our request – or maybe just mine -, Ollie provided a great introductory talk to meteorology and its basic laws to help us understand how everything in meteorology is interconnected; thus also connected with infrasound. Jeff also gave an impressive introduction to hearing, and while it is hard to keep track of all the technical vocabulary, we enjoyed the presentation about auditory models and the anatomical comparison between terrestrial and aquatic bird species (read more about his research on this post). As I wanted to get more involved with the weather data analysis, Ollie shared a tutorial and gave me a script to download the ECMWF data, and we chatted about the links between atmospheric pressure and wind. It was all very helpful and I hope I can go visit him and Jelle at the Royal Netherlands Meteorological Institute soon!

During the second week, I focused more on working with Tommy to assess the effect of wind in albatross behavior during foraging trips via Hidden Markov modeling.

tommy-track.jpeg
Albatross track with inferred behaviors

There are at least two packages in R to implement Hidden Markov models (moveHMM and momentuHMM), making our tasks easier, but knowing how to parametrize the models and correctly interpret the results require understanding the intricacies of the methods as well as the ecology of the species. Together, a statistician and an ecologist, we’re trying to bring both components to the table for this piece of research—a manuscript should be ready in the next few months!

Working in Sam’s lab for a couple of weeks, I had the opportunity to participate in weekly lab meetings to discuss scientific articles and get to know a bit more about their research. I really felt like I was part of the lab during my stay, which is always a plus.

Avignon

After a long trip from Liverpool that took me the whole day, I arrived to Avignon for a workshop on SPDE (Stochastic Partial Differential Equations) modeling and INLA (Integrated Nested Laplace Approximation). I think that there is no need to say that this was a different scientific community than the seabird ecology department at the University of Liverpool.

inla
D. Allard, from RESSTE, introducing the SPDE workshop (picture taken by O. Gimenez)

The workshop was organized by the RESeau Statistiques pour données Spatio-TEmporelles (RESSTE) network and gathered a diverse public between applied (but quantitative) and theoretical researchers. Since I am considering using continuous time models for movement further in the HFSP project, I thought that learning an INLA approach would be handy for this. INLA can be a great solution to estimate parameters of complex models, and the organizers of the workshop provided a short introduction on how to express statistical models as SPDE that could be solved by INLA. Then, they showed the implementation of INLA in R (using R-INLA) with some examples of time series and spatial point processes. Some coauthors of the R-INLA and INLABRU packages were present and contributed to the seminars with their comments; that also made me realized that implementing INLA can be an incredibly difficult task for a model that has not been considered yet in the package. The seminars were complemented by the series of presentations from several participants on their research applying INLA.

Nantes

nantes-vue-de-mon-appart
View from my old flat in Nantes. I miss that.

After Avignon, I spent the weekend with some friends in Sète. Sète is a lovely small fishing city on the Mediterranean and one of the places I stayed in during my PhD. The weekend in Sète gave way to a three-week stay in Nantes. I did a previous postdoc at IFREMER in Nantes, and I was coming back this time to work on ‘old’ papers: one reviewing metrics to assess dyadic joint movement (now published in Movement Ecology) and one applying these metrics to several fleet movement data and learning about their collective behavior (almost ready to submit). My collaborators, which are not all from IFREMER, are quantitative researchers and founders of the trajectometry group Path Tool and analysIS (PathTIS). During my visit, we programmed meetings not only to discuss about the papers we needed to finish), but also to talk about my future projects to come again and collaborate on movement modeling papers.

Nantes was the last destination of this trip before heading back to Florida. All in all, I am very happy with all the work done and the scientific discussions. …And the wine, the food, the pubs, the funny conversations, the walks under the rain and the music. Thanks to everybody who welcomed me into their labs and cities, and the friends who hosted me. I had a blast and I hope to see you soon!

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s