LAB 1 - Introduction to using GIS tools, accessing occurrence data, georeferencing.

PART 1.  GIS SKILL UP

1.         Download DIVA-GIS (sorry it is just for PCs. There is a rumor of a Java version but nothing more)

2.          Download the tutorial and accompanying data for the tutorial.  Please run through the tutorial.

3.         Download worldclim data here:  http://www.diva-gis.org/climate.htm   start with the lowest resolution data to start.

You can load worldclim data into DIVA-GIS by going to the Tools->Options->Climate and pointing to the location of the worldclim data.  Do so.

To visualize climate data, go to Data->Climate->Map and choose a climate variable to open, and extent.  This should show you a “climate map”.

4.         Download all sorts of relevant data here: http://biogeo.berkeley.edu/bgm/gdata.php

            I haven't used this new download capacity yet, but it looks REALLY good.

Try it out.  To do so, download the stuff you want and keep in mind there are two formats for  dowloaded data:  raster data and shape file data.  Raster data are "GIS layers" that are represented as grids (equally sized squares with each square having a value representing some environmental measurement eg. mean annual temperature).  Shape files more typically contain geospatial “footprints” like lakes shape and size, which also may have values associated with them.  One _can_ convert a raster file into a shape file and vice versa (with some caveats).

DIVA-GIS has some excellent import functions for raster data in formats that are not native. Raster file format conversions are done under the Data -> Import to Grid File menu.

  1. Upload point-data species occurrences into DIVA using the Data -> Import points to shapefile menu.  See below for how to get your own species occurrence dataset from GBIF, MaNIS, HerpNET, Antweb, ORNIS, FISHNET, etc.  After you have acquired a dataset, open it in Excel and save it is a tab-delimited text file.  Make sure to close the Excel file (DIVA gets confused otherwise).  Now give the input file name location (easiest to browse for it) and decide on an output file name.  Check the part of the menu called “Field Delimiter” and make sure the radio button “Tab” is checked.  If DIVA doesn’t automatically detect the longitude and latitude fields in the species occurrence dataset you uploaded, set these yourself using the drop down menus.  Once all looks good, hit import.  You should see the points on whatever basemap you have selected.
  2. Try using the Modeling -> BioClim/Domain functions, except don’t do “Predict” (bioclim/domain modeling is a bad idea … more next week).  The other tabs are worth exploring; they give some context about the frequency and distribution of environments where your samples were “collected”.  They should be somewhat self-explanatory.
  3. One very useful function in DIVA-GIS is “environmental extraction”.  That is, you can determine underlying environmental data for any point on a map.  So, for example, you can determine elevation, mean annual temperature, precipitation, etc etc. for any point occurrence.  To do this, go to Data -> Extract Values By Points.  You can extract values from the climate data you set up earlier (select From Climate Data).  If you select the Bioclim check-box, you get eighteen bioclim variables outputted (see below for Bioclim variables).  You can also select other grids or stacks of interest and extract values from them as well. 

The file that is saved is a tab-delimited text file that can be opened in Excel. 

  1. A stack is a set of grids that are layered ontop of one another.  Stacks are very useful if you want to do calculations across multiple grids or extract multiple values from a set of related layers.  You can make a stack from a set of grids using the “Stack -> Make Stack” function.  We will be using DIVA-GIS a  lot to look at niche modeling results, so more on a lot of functions then.

 

BONUS FUN STUFF:  Try out the point to grid function under the Analysis menu.  See how those works.  Try running a BioClim model JUST FOR FUN (but remember, this is not a good modeling suite).  I have tried using the spatial autocorrelation analyses in DIVA but they seem memory limited in a lot of cases.  We will do more in Spatial Analysis in Macroecology than in DIVA, but it is a good chance to begin to “tinker”.

 

BIOCLIM VARIABLES: (BIO1 = Annual Mean Temperature, BIO2 = Mean Diurnal Range (Mean of monthly (max temp - min temp)), BIO3 = Isothermality (P2/P7) (* 100), BIO4 = Temperature Seasonality (standard deviation *100), BIO5 = Max Temperature of Warmest Month, BIO6 = Min Temperature of Coldest Month, BIO7 = Temperature Annual Range (P5-P6), BIO8 = Mean Temperature of Wettest Quarter , BIO9 = Mean Temperature of Driest Quarter, BIO10 = Mean Temperature of Warmest Quarter, BIO11 = Mean Temperature of Coldest Quarter, BIO12 = Annual Precipitation, BIO13 = Precipitation of Wettest Month, BIO14 = Precipitation of Driest Month, BIO15 = Precipitation Seasonality (Coefficient of Variation), BIO16 = Precipitation of Wettest Quarter, BIO17 = Precipitation of Driest Quarter, BIO18 = Precipitation of Warmest Quarter, BIO19 = Precipitation of Coldest Quarter. 

 

PART II.   Acquiring Species Occurrence Data

  1. Go to http://data.gbif.org/
    1. What is GBIF?  It is a global organization that is helping to federate species occurrence information (when and where species occur). 
    2. GBIF is a “presence only” dataset, but it is huge.  150 million records.
    3. All participating organizations use the same data standard and transmission protocols.
    4. The standard format is called DarwinCore. 
    5. More here: http://wiki.tdwg.org/twiki/bin/view/DarwinCore/WebHome
    6. GBIF also facilitates training and building biodiversity capacity.
  2. Select “Explore Species” and type a taxonomic name (probably good to try a genus or species name) into the text box. 
  3. Once results are returned, try “Matching Records on Map”.  Note that the view is coarse (1degX1deg blocks).
  4. Next download spreadsheet of results.  Note this takes time.  Just do a default download of all fields.
  5. Save zip file of occurrences to your local drive
  6. Unzip file.  Open the occurrence “spreadsheet” in Excel.  Note the format is tab-delimited text.
  7. Take a look at the field formats. 
  8. Sort your records in Excel by latitude.
  9. How many have latitude and longitude?
  10.  Pay special attention to the field next to latitude and longitude called coordinate uncertainty.  How many records have coordinate uncertainty?
  11. Coordinate uncertainty expresses the “accuracy and precision” of georeferencing.  Most records in GBIF are legacy records before GPS units were commonplace.  Even GPS units have error… however they are generally more precise than locality precision. 
  12. Retrospective georeferencing is a way to assign geospatial coordinates to a locality desc.
  13. More below on georeferencing.
  14. Records without coordinate uncertainty are not fully trustworthy for use.  You don’t know how “good” they are.
  15. Try to delete all “superfluous” columns, leaving at minimum taxon, longitude, latitude.  Load this into DIVA-GIS.  It is that easy.

 

BIOGEOMANCER:

 

            The last task of the day is to retrospectively georeference some of the records from GBIF that lack geospatial coordinate and/or coordinate precision.

 

  1. Go to http://bg.berkeley.edu/latest/
  2. Read the help (if you have complaints, let me know…I wrote it).
  3. Try georeferencing a locality
  4. Try editing location and uncertainty amount.