Re the graph. It was provided to give people an idea of the number of records in the database for each month, something I hope most would have got from it. The graph and my text in no way suggested that there was any relationship to the number of individual birds present and it is inappropriate to try to draw these conclusions from the statistics provided. Care should be taken when considering the periods with no records – it does not mean the species is not present, just that there are no records in the database to indicate it is present.
I will continue to provide information to the chatline. I hope that people do not extend it to inappropriate conclusions but know that some will and there is no way that this can be avoided. I do not want to have to provide a full quality statement for these data as, in my experience, very few people read or understand these.
The COG Data License Agreement signed by people using the COG data contains the following:
“The data is provided with the following statement as to its quality, which must accompany the data in any report, document or publication in which the data is used.
This data was collected by amateur bird watchers on general surveys. Note that on some occasions more than one person may have recorded bird sightings on the same day.”
While this statement does not provide a complete quality statement, it does try to state that these data do, in some cases, contain multiple records for the same sighting.
Steve
From: Philip Veerman [
Sent: Friday, 13 June 2014 1:13 PM
To: 'Geoffrey Dabb';
Subject: RE: [canberrabirds] Little Bittern
Thanks Geoff, Sure it is great that we have the resources to do such a thing (built up over quite some time to create the systems to do it quickly) and that Steve put in the effort to provide the graph. I just wanted to highlight that the "surely" in your sentences actually does occur: "Of course the data compilation is largely unstructured and subject to all kinds of biases. Surely we take those into account in making use of them." Because given a graph like that, it is easy sometimes not to do so and may give a wrong impression. I am not suggesting it is wrong in this case. It does help to actually point out the possible biases.
As for: Now Philip questions the value of counting records month by month. In this case it would help to give a count of one record only to each known or suspected single occurrence. Of course that is more difficult and also involves assumptions and a question of what to do if one bird is seen most days of one month and then into the next month........ The GBS has similar issues on a week scale and The GBS Report (and the earlier COG ABR) do to some extent, by giving monthly based abundance graphs. I would also say it is somehow more significant if a vagrant individual does stay for a long time, rather than just appear on one day. (Like the White-fronted Honeyeater that was in my yard every day for nine weeks.)
-----Original Message-----From: Geoffrey Dabb Sent: Friday, 13 June 2014 8:01 AM To: m("canberrabirds.org.au","canberrabirds");">
Subject: FW: [canberrabirds] Little Bittern
This seems to me to be confusing the value of Steve’s kind of graph with the deeper question: ‘What do COG’s records really show?’ - ie what limitations should be taken into account in their interpretation. Of course the data compilation is largely unstructured and subject to all kinds of biases. Surely we take those into account in making use of them.
On the face of it, there is strong evidence the Little Bittern is present here mainly in the warmer months - although the recent sightings show that is not an invariable rule. What is the true underlying pattern of movement? - as for so many Australian species we do not know. For many species, much of the reporting cited in HANZAB under ‘Movements’ points in different directions.
Now Philip questions the value of counting records month by month - as in the graph. The graph is simply a quick way of showing what the records tell us about reports over a year. These quick snapshots by Martin and now Steve have added some interest to this chatline, and I for one appreciate them. If (subject to Steve’s patience and goodwill) you wish to go deeper you can investigate individual reports and their exact time and location - gain an idea of their reliability, even. This will show whether a cluster of reports in November are attributable to a particular year and to one or a series of reporters, and – with a high degree of probability - whether they relate to a single appearance, possibly whether more than a single bird was involved.
In short, the graph does not represent all the data held by COG. It merely offers one way of looking at it. The question ‘How real are these data in terms of the species presence?’ is not one that should be directed to the graph.
I am aware Philip knows all this, and merely lacked his usual clarity in _expression_.
Maybe Mark is commenting on the overall level of records of the species in our AOI. That is fine. I vaguely recall his story about When COG was surveying for the then proposed site for the National Museum at Yarramundi Reach, which I assisted in, ...... as I was involved in that too. I was commenting on the issue about whether the implication to be taken from Duncan's graph is real (as distinct from statistically correct) in suggesting there are more Little Bitterns here in the warmer months. The aspect of being cryptic is very relevant in the situation of a species being present in similar numbers all year but by its behaviour is well hidden for part of the year and easily observed at other times. I don't know if that is relevant for this species. But it could be and thus is a point I overlooked. It could also be that people are more willing to proceed to wade chest deep through the reeds during warm weather than cold, and that could influence recording rates too.
As Martin says, chatline mentions are not official records but they are a quick way of establishing a pattern of dates. Also it is not needed for each report (especially of repeats) to make it to the Rarities Panel to be officially on the record. Although them having several is better than none or one if helps or it is important enough.
-----Original Message-----
From: Mark Clayton [m("bigpond.com","chollop7");">]
Sent: Thursday, 12 June 2014 2:44 PM
To: 'Philip Veerman'; 'Wallaces'; m("canberrabirds.org.au","canberrabirds");">
Subject: RE: [canberrabirds] Little Bittern
One of the big problems with cryptic species is that they are just that – cryptic! When COG was surveying for the then proposed site for the National Museum at Yarramundi Reach, which I assisted in, I asked if anyone bothered to check the reed beds for Little (now Black-backed) Bitterns. When told “NO” I proceeded to wade chest deep through the reeds and successfully flushed a couple. Since then there have been quite a few records, often by people using canoes to scan the lake side edges of the reeds. Looking at Lake Ginninderra and the creek that flows into it (past the Crace and Giralang ponds and close to the large McKellar pond where the two current species are being recorded), there are large expanses of reeds that could easily hide both species of bittern and that theoretically are inaccessible except y boat.
I don’t know about the south side of town but in the newer Gungahlin suburbs there are numerous small reed and rush-lined ponds that would be quite capable of hosting both bittern species. How many people visit these on a regular basis? I don’t always agree with Philip but I think he is right in this case, with most of the records being of the same birds being constantly reported by people seeing them for the first time and putting them on the chat line. How many make it to the Rarities Panel I don’t know but all records of these species should go through the panel to make sure they are officially recorded as well as putting in something like an “incidental report form” which can be found on the COG website.
Mark
Thanks for that. So mainly a warm month visitor. (The first RAOU Atlas suggests that too, the second RAOU Atlas is hard to perceive any trend). So is the current winter record record unusual? It does beg the question though about counting the number of records by month. Are these separate records? How real are these data in terms of the species presence? In this species I perceive that there would be very few separate events. The species is rarely here and when here many people go to find it. So there may well be a lot of recounting of the same birds in this graph. Which can make a trend look bigger than it should.
-----Original Message----From: Wallaces [m("bigpond.net.au","skcbf");">] Sent: Thursday, 12 June 2014 12:48 PM To: 'canberrabirds' Subject: RE: [canberrabirds] Little Bittern
Previous records of Australian Little Bittern are all from October to March. The graph below shows the number of records by month for the period 1 July 1981 to 30 June 2013.
Steve
