PPGM

THIS PAGE CURRENTLY UNDER CONSTRUCTION. SOME SECTIONS MAY BE INCOMPLETE. THANK YOU FOR YOUR PATIENCE WHILE WE WORK ON THIS.

PaleoPhyloGeographic Models are a method conceptualized by Dr Michelle Lawing and implemented in various publications. ppgm is now available on CRAN, and can be downloaded from R using the following code: install.packages("ppgm"). This package can also be directly installed from GitHub, which will have more recent bug fixes (see documentation).

Here we provide some extra help files for users wishing to implement PPGMs using our published R package ppgm. If you use the ppgm package in a publication, please let us know and we will add you to our list.

If you have any questions not directly addressed on this page, please email alexandra.howard@ag.tamu.edu and I will endeavor to respond quickly.

Topics

How To Gather and Format Extant Occurrences

For your extant occurrences, we recommend gathering data from an online repository such as GBIF (Global Biodiversity Information Facility). Information about GBIF can be found here.

Click OCCURRENCES to examine data

If your search matches a taxonomic ID, click YES here

Dataset types in GBIF

When selecting occurrences, we recommend iNaturalist Research Grade observations as a reliable source of information for the modern day distribution of species. This can be selected in the GBIF user interface.

iNaturalist is an online repository of crowdsourced species identifications. Research-grade observations pass several qualifying factors such as specific date, time, location data, as well as ID supported by two or more independent observers, with photo evidence of a wild animal. Research-grade iNaturalist observations are commonly used in the biological and conservation sciences.

Occurrences downloaded from GBIF.org comes with a lot of associated data that is not needed for running a PPGM. Extract the following columns from the dataset: decimalLatitude, decimalLongitude, and specificEpithet

For ppgm, the columns must be in the following order: Species (specificEpithet), Longitude (decimalLongitude), Latitude (decimalLatitude). Check the order! This is the reverse of the order that defaults from GBIF.


How to Gather and Format Fossil Occurrences

Fossil occurrences can be used from any source as long as there is associated paleolongitude and paleolatitude data with each occurrence. We also caution that the further each fossil is from associated paleoclimatic data, the less specific the results will be. Fossils can either be from known locations in the tree or unknown, which can be specified later within the ppgm workflow.

Alternatively, fossil occurrences can be downloaded from the Paleobiology Database. The PBDB is a public database for paleontological data from scientific publications, uploaded by contributing scientists. Remember to cite the PBDB in any published manuscript using data from this source.

Remember to keep your original download from the PBDB, as this includes useful information about each fossil occurrences such as the original publication, stratigraphic and lithologic information, and country of origin.

For ppgm we need the columns max_ma, min_ma, paleolng, and paleolat. For fossils from relatively recent deposits (eg. Pleistocene), lng and lat can be used instead of paleolng and paleolat. This is because the paleolng and paleolat represent the position of the land during the time of fossil occurrence, which may be different to current lng and lat due to continental shift through time. Young deposits such as Pleistocene do not have this data due to the continents being in their modern positions at this time period.

Do not include multiple observations of fossils from single localities (long/lat). The package does not take any abundance into account, so multiple fossils from the same locality will only cause problems in the code. If you are having errors, try removing the number of fossils that occur in the same time period.

For input into ppgm the data must be in the following order: MinAge, MaxAge, Longitude, Latitude. If wanting to include fossils on specific edges of the tree, these must be stored as a separate vector.

Caution: We advise against too many fossils in the same time period on the same branch, as this will cause a break in the code. When adding fossils the function addFossil will randomly add the fossils iteratively to the tree according to branches that occur at the age of the fossil. If many fossils are added at the same time on the same branch, the branches will get shorter and shorter and likely cause an error.


Publications Using PPGM

  • Rivera et al (2020) Reconstructing historical shifts in suitable habitat of Sceloporus lineages using phylogenetic niche modelling. Journal of Biogeography 47(10): 2117-2128
  • Lawing et al (2016) Including Fossils in Phylogenetic Climate Reconstructions: A Deep Time Perspective on the Climatic Niche Evolution and Diversification of Spiny Lizards (Sceloporus). The American Naturalist 188(2): 133-148
  • Rödder et al (2013) Evaluating the Significance of Paleophylogeographic Species Distribution Models in Reconstructing Quaternary Range Shifts of Neartic Chelonians. PLoS ONE 8(10): e72855
  • Lawing & Polly (2011) Pleistocene Climate, Phylogeny, and Climate Envelope Models: An Integrative Approach to Better Understand Species’ Response to Climate Change. PLoS ONE 6(12): e28554

FAQ’s