rbatt / trawldiversity

Long-term trends in regional species richness

trawlDiversity

Gradual changes in range size accompany long-term trends in species richness

Authors Ryan D. Batt, James W. Morley, Rebecca L. Selden, Morgan W. Tingley, Malin L. Pinsky

This R package is designed to reproduce the statistical analysis and visualization of a substantial portion of my postdoctoral research, which is in progress.

trawlDiversity relies on the trawlData R package. trawlData contains scientific bottom trawl surveys of marine animals all around the North American coastline during the past several decades.

Interestingly, many of these species are known to be shifting their geographic range in response to climate change (see OceanAdapt Website and OceanAdapt Repo). However, what does this mean for groups of interacting species?

My goal is to understand how the players in these food webs may have changed -- have species been gained, or lost? If species richness has changed, has it happened with gradual or rapid shifts in the ranges of individual species? The answers to these questions will help us understand how these systems have changed in the past, and how marine ecosystems may respond to environmental change in the future.

A centerpiece of this analysis is the use of a multispecies occupancy model (MSOM) to estimate species richness while accounting for imperfect detection of species by the sampling methods use in the surveys (this is a common problem with any biodiversity study). These corrections have never been made in marine studies before, but can crucially effect estimates of biodiversity.

The MSOM models run in Stan or JAGS. These are additional programs that won't be installed directly through R. Furthermore, the models are computationally quite intensive. For this reason, this repository not only contains final results of the analysis, but also contains intermediate results.

https://onlinelibrary.wiley.com/doi/abs/10.1111/ele.12812

trawldiversity's Issues

November Revision to-do

• Categories and range in richness/ initial-final richness. Can "colonizers" explain changes in richness?

• As suggested by @mtingley, we should report (1) the average number of tows per site per year, and the maximum number of tows per site per year. (2) The number of sites per region (ranges)

• For which regions did we use truncated time series? In most regions I believe the sampling frequency and location was the most important factor for choosing to drop years. In other regions it may have been the large number of initial colonizations, e.g. EBS.

• Define how day of year was adjusted for each region, and pair with density plots for number of hauls (y axis) vs day of year, and a different line for each year

• List which species were removed, and why?

• Effect on bin size (lon-lat and depth) on number of sites sampled

can't open .RData files in cloned repo

Hi Ryan! I forked and cloned this repo, but when I go to open any of the data/*.Rdata files, I get the following error:

> load("~/github/trawlDiversity/data/spp_master.RData")
Error: bad restore file magic number (file may be corrupted) -- no data loaded
In addition: Warning message:
file ‘spp_master.RData’ has magic number 'versi'
  Use of save versions prior to 2 is deprecated 

What's the best way to access these files?

Thanks so much for making all of these resources open access!

remove leopard skate?

http://www.iucnredlist.org/details/80671020/0

Apparently it wasn't declared its own species until ~2010; that's about when it shows up in the data.