Wednesday 4 January 2012

Atlas of Living Australia: Where the woozle wasn't (and was)

I’ve been playing around with the Atlas of Living Australia to explore how useful it might be in undertaking biogeographic studies from the comfort of my frayed but functional ergonomic chair. The ALA is an interactive online database of Australia’s flora and fauna. Its data comes from a variety of sources, mostly museum and herbarium collections (supported by specimens), but also observations submitted by the public (sometimes supported by photographs). Because the usefulness of the Atlas depends on the comprehensiveness and accuracy of its data, some of the records are bound to be dodgy. The ALA is working on improving the quality of harvested data, although there isn’t a lot that can be done at that end about misidentifications. So, with that caveat, I had a wander through the database. Why not wander through it with me and see what you think.

I picked Lumholtz’s tree kangaroo (Dendrolagus lumholtzi) as a candidate and had a look at what the Atlas has to offer on this species.

Searching on the name gives some basic information on the taxon, including an illustration.


Overview provides a map of the distribution based on records. Depending on your requirements, this might be sufficient. If not, you can examine the records as a list or plotted on an interactive map. Overview also offers information about conservation status (symbols on right).


You can also get a full list of taxa with the native ones all dressed up for Australia Day.


The source of (some of) the records gives a rough indication of their reliability.


And their distribution through time would probably be useful for investigating migratory or seasonal species if there were enough data.


The datasets come with a number of filters by which you could sift data (continued over two images).


I sifted the Dendrolagus lumholtzi records by precision of locality — between 1 and 5 km. The map looked mostly reasonable, although there were a handful of potentially dodgy records.


Tarzali is a known spot for tree roos, so I checked out that record.


The data came from CSIRO's Australian National Wildlife Collection and had a registration number attached to it. According to the details on the pop up, the specimen came from Hogan's Road. Entirely plausible.


Somewhat less plausible is this record from west of Mareeba. I've switched to the satellite view to give you an idea of the country out that way. Not typical tree roo habitat.


When I looked at the details, it was obvious that the coordinates are wrong. This specimen came from woodland SW of Atherton.


A startling cluster of records — seven tree roos apparently from what's not quite the middle of nowhere, but somewhere near it. These data must be viewed as being Probably Not All That Reliable. Still, if I really were doing a study of tree roos, I'd shoot an email down to the SAM with rego numbers and find out the story. Dodgy coordinates? Fossil material? Escapees from a travelling marsupial circus?


It is also possible to sift out type material. This is the holotype for Dendrolagus lumholtzi. So, much fun to be had with a few clicks of the mouse.


You can have even more fun by looking up the records for your immediate vicinity. When I checked my address on the database, I was surprised to find only two species of frog recorded within 5 km. Obviously an underestimate, because I can hear three species calling as I type this, and I know there are several others around. So I know what I have to do — upload the data.


I haven't fully explored the Atlas. In fact, I haven't even scratched the surface. It even allows you to map distributions in relation to environmental variables such as climate factors and vegetation. Hours of entertainment.

I have yet to decide whether it will be sufficient for some office chair research (more businesslike than armchair research), but it does have a lot of potential. And did I mention the hours of entertainment?