Tuesday, 23 March 2010

The birds of East South(ish) Canberra

To begin with, let me explain the title of this post.  South Canberra is an official designation of the area of the urban part of the ACT to the South of Lake Burley Griffin and bounded by a range of low hills, separating this township - not at all like the SOuth WEst TOwnships of Johannesburg - from those of Weston Creek and Woden.    The study is largely concerned with the Eastern component of South Canberra, but also strays across the Molonglo into North Canberra: hence the (ish).

I have recently been looking at the implications of the grid of latitudes and longitudes used by the Canberra Ornithologists Group (COG) to summarise observations made by members.  One of the technical issues raised in considering the usefulness of such grids is the extent to which an observation from a site within a grid cell can be considered representative of the cell as a whole.

Obviously this depends on the homogeneity of the habitats across the cell.  In some cases the cell will contain a single type of habitat (at least at a broad level): by way of example our block is within cell Q16 which is reasonably consistent as a rural residential area, with some variation in density of the remnant tree cover.  In other cases however the habitats within the grids are remarkably heterogeneous.  Cell L14 is a good example of the latter situation, containing some suburbs (about 75% of the area of the cell) and an area of water features and pastures around the Molonglo River (the remaining 25%).  [There is also some variability in the types of suburbs including some of the most up-market suburbs in Canberra; some lower-rent residential areas; and business oriented areas around the Parliamentary area.  However at this stage I am not going to obsess about the latter.]

The other attribute of Grid L14 is that it is usually one of the areas with the greatest number of data reports.  It is probable that this reflects the relatively high diversity of the bird life in the area which in turn could be attributable to the mixture of habitats. 

The catalyst for undertaking this report was noting that for 2009 all the reports of General Observations for cell L14 came from the damp 25% of the cell.  Put simply, the data for the cell are unlikely to be (I am tempted to say 'not' ) representative of the cell as a whole.   This becomes important when the data is being analysed by policy makers: by way of example issues recently arose about management of the Urban Forest (the street and garden trees of the suburbs) within the area covered by the cell.  Without a great deal of care the data for cell L14 could be highly misleading if used in that context.  So I decided to investigate what birds could be found in the more typical suburban element of L14 and the investigation I commenced is the main thrust of this post.

Before getting to that exercise I will note that COG also undertakes a Garden Bird Survey (GBS) of which the author is the Coordinator.  The GBS has been running for 28 years and over that time 11 sites have operated in cell L14 for one or more years.  The location of these sites is shown as a white star in the attached graphic: those in the Parkes/Barton area are mainly sites people operated from their workplaces. (The yellow pin marked 'Kelly's shows the focus of the wetlands.)

Over the period of the GBS these sites have reported 116 species at least once. This ranks it 17th of the 41 cells with at least one GBS report (the highest diversity is cell J14 with 162 species reflecting a wide range of habitats and expert observers.).  So in terms of a list of possible species observed in the area, not all is lost.  However the GBS methodology is not easy to link with the general observations.

The enquiry which I am mainly writing about here consisted of choosing 6 points across the suburban component of grid L14 and undertaking a 20 minute survey of the birds in 2 hectare sites based on those points.  The selection of the sites was purposive in that I wanted to cover:
  • the entire suburban part of the cell (thus a site in Duntroon as well as the 5 to the South of the river);
  • well to do areas and less economically well off;
  • office dominated areas as well as residential areas; and
  • a number of different suburbs.

However I did want to find some birds so  introduced a bias by looking for areas which included open space rather than absolutely built-over areas.  I also avoided a few areas (eg the Russell Offices, home to Australia's intelligence community, where someone with binoculars and a camera - even wearing an Akubra rather than a keffiyeh - may be looked upon with suspicion).

This image shows the location of the 6 sites selected marked with a red numbered 'teardrop'.

I undertook the field work for the exercise on 22 March 2010 starting (for reasons of logistical convenience) at site 3 at 9:35am.  The weather was fine and mild: little wind or cloud and the temperature about 20 - 24 C.  Including travelling time the process took 2 hours 40 from start to finish (quite good efficiency with 6x20 minutes  = 2 hours or 60% of the time actually looking for birds). 

Over the whole period I observed (including a few 'heard-only' records) 24 species.  Twelve species were only recorded at 1 site while three species (Sulphur-crested Cockatoo, Australian Magpie and Pied Currawong) were found at all 6 sites.  Some general notes about my expectations and analysis follow the comments in the next section about the various sites.

Site 1: Rose Gardens, Old Parliament House, Parkes.  This site was intended to represent an open area as well as this 'suburb'.  8 species were observed, of which Gang-gang Cockatoo and White-winged Chough were not seen anywhere else.  Although it was hoped that the presence of flowers would attract species feeding on nectar and insects this did not happen.
Site 2 Blackall St, Barton: Representative of the office-building environment which occupies a significant proportion of the area of the cell.  8 species with Superb Fairy-wren and Willie Wagtail (surprisingly) only reported at this site.While the office and car parks areas were almost completely devoid of bird-life the gardens along the Northern side of Blackall Street were relatively bird rich, probably reflecting the density of the vegetation.
Site 3: General Bridges grave, Duntroon:  This is a low peak in bushland above the Royal Military College.  It is intended to both get a site on the Northside and to represent the remnants of bushland throughout the cell.  It was the first site surveyed as it was easiest to get to at 9:30.  6 species were observed.  This was the lowest number for any of the 6 sites and possibly reflects the rather degraded state of the under-storey which appeared to have been burnt in the recent past.  The two species unique to the site were Laughing Kookaburra and Rufous Whistler.
Site 4: Bass Gardens, Griffith:  This is an older developed area of Canberra and includes a large park with many conifers and some large eucalypts.  9 species were recorded here with several (notably 9 Crested Pigeons and at least 12 Pied Currawongs) in relatively large numbers.  Much excitement - to the observer as well as the avifauna - was generated by a Brown Goshawk launching an attack from a concealed perch in a spruce on a group of Crested Pigeons feeding on the ground.   The call of a Satin Bowerbird was heard but the bird could not be located.
Site 5: Lefroy St Griffiths.  Another older area with large trees and wide nature strips.  It was adjacent to St Edmunds College.  12 species were recorded   Uniques were Australian King-parrot (heard only, probably feednig on the acorns common in the area); Spotted Pardalote; and Australian Raven (6 feeding on some garbage adjacent to the school grounds).
Site 6: Nimbin St, Narrabundah.  A loop street with a grassy area on the NE and NW outside of the houses.  The houses were relatively small and most trees were exotic species.  Again 12 species were recorded.  Uniques were Eastern Rosellas  feeding in a grassy area behind the houses A pleasant sight here was a Pied Currawong attempting to fly off carrying a medium-sized rat which it had found (and, from the appearance of the rodent's carcase, killed recently).

"Missing in action"
I was somewhat surprised that I did not record a single House Sparrow at any site.  It was also surprising, but in a pleasant way, that Common Myna were only at 2 sites.

Very few small passerines were seen at any site.  I expect this reflects the paucity of food resources available for them.  The only site at which a significant amount of plants were in flower was the rose garden and I suspect that insects are not encouraged there thus deterring pure insectivores (wagtails and flycatchers for example) and omnivores (honeyeaters).

This was one survey, undertaken by an observer unfamiliar with the area.  It is thus quite likely that some species were missed.  Also, in many parts of the sites much of the area was behind opaque fences and thus all that could be seen were the birds in the front yards and on powerlines etc.

As noted above there was very little blossom evident in the gardens and most of the street trees were exotic.

It might also be expected that a number of migrant species that are observed in the area - Cuckoos and flycatchers being notable examples - have already moved out.

Given the caveats above it is dangerous to make a great deal of this initial data set.

It is however clear that the range of bird species seen were very different from those that could be expected in the wetlands around the Molonglo.  It is thus concluded that this exercise has proven the a priori expectation that the "normal" run of COG General Records do a poor job of representing the birdlife of this area.

It is also noted that the species observed differed significantly between sites.  It is possible that additional observations would reduce this variability and that use of the relevant GBS data may go some way towards explaining it. 

A key point which does emerge is that 5 of the species observed were hollow-nesting parrots and thus these data might have made some (very) small contribution to Urban Forest issues which the Wetlands focussed data could not.

A final comment is that even this work omitted some suburbs from the sample.  Nothing was done for Kingston, Red Hill, Fyshwick, Campbell or Russell all of which were, in part at least, in the cell.  Possibly the sites included  could represent them  -for example site 3 could represent Russell as well as Duntroon and site 4 could represent Red Hill - but this cannot be explored withe data as it exists.

To the future
A key attribute of the 2Ha, 20 minute survey approach is that the surveys should be able to be replicated such that a time series is built up.  It would be surprising, to say the least if a repeat of this exercise in  September or October did not produce a different, and much longer, species list.

It is also evident from the GBS list that a number of species (particularly waterbirds and raptors) will be observed through a principle of serendipity.  Indeed, sighting the Brown Goshawk in site 4 could be seen as an example of this.  Again further observations will add to the list.  Thus I intend to repeat these surveys on at least a bi-annual basis.

No comments: