Indicator Birds

 I recently asked a question on Mallacoota Birds FB Group about which birds were the best indicator that an area had been adequately surveyed.  By the phrase "adequately surveyed" I mean that the list of birds available for the area represents the set of bird species that are actually there.  Getting more involved in the meaning, in this case it is more along the lines of "proof by exception": if the indicator species isn't recorded then the list is inadequate.

2 responses to my question suggested New Holland Honeyeater was an indicator while another examined Birdata submissions to show Superb Fairy-wren recorded in 2,124 or 47.03% of total surveys followed by Grey Fantail 39.19% and Red Wattlebird 39.04%. NHH comes in at 35.34%.  That response also suggested eBird might have similar information.

The problem with data 'as reported' from both those sources is that the sites used are very much determined by how the observers decide to report.  It can mean that: 

• there are lots of sites in close proximity (eg there are at least 20 locations in eBird within the Foreshore Campground); or 
• people do a single list for a visit probably covering many areas (eg 11.4% of eBird lists are under a locality simply called Mallacoota).

The approach in Tanzania (as was the case for the Atlases of Australian Birds) was based on grid cells defined by lines of latitude and longitude.  Normally I fulminate against such boundaries as ignoring anything to do with biology but in this case they suit my purpose, so are given a pass!  I have defined a grid based upon 0.1 degrees which for my definition of the Mallacoota District gives the following set of cells.  (Note that due to my poor drawing skills the SW corner of the polygon shown below is not quite precisely represented: it is actually 37.4S 149.5E which includes the Wingan Inlet campsite.)

I classified each eBird locality (I don't have ready access to the Birdata set, but would expect a similar result)  to a cell as defined by its geocoordinates.  14 have some land in them, 15 have some ocean in them.  6 have both land and ocean.  The simplest indicator is to look at the number of localities defined for each cell.

It is immediately suggested that the Wingan area (cells A2 to A4) and the far East of the Howe area (cells E1 and E2) have had little attention.

Looking at the number of species reported ameliorates that conclusion with the species count suggesting that when people have visited those areas they have found quite a lot of birds.  

Cells A2 and A3 clearly need more attention.  Offshore there have been a few pelagic trips (and intentions to have more, during the 2 Big Weekends have been ruled out by weather conditions).

I have now looked at the species suggested by others.  For land sites Grey Fantail has been reported from all 13 of the land sites with an eBird Location.

New Holland Honeyeater has been reported from 12 sites - missed in cell A2.

Superb Fairywren has also been reported from 12 sites, with e2 being missed.

For sea birds Crested Tern has the greatest coverage with 13 cells represented.  The 2 cells 'close in' without Crested Tern records have not been eBirded.

The other contender for indicative seabird is Silver Gull but that has - to some extent surprisingly - not been covered from the pelagic cells with a total of only 11 cells covered.

My primary conclusion is thus that to consider we have fair coverage of land birds we should have a record in a cell for Grey Fantail, New Holland Honeyeater and Superb Fairywren.  For Seabirds, Crested Tern is the major indicator.

My second conclusion, linking to the reason Tanzania had the concept of an indicator bird, is that some birding needs to done along the roads to Wingan Inlet, especially East Wingan Rd which runs down the middle of cell A3!