|
Hi All,
The GBS is an important data source and it matters that
interpretation is properly understood. The graph that Martin provided presumably
shows the number of breeding observations combined, of all types, by
week. That has peripheral usefulness. However the benefit of the GBS Report
graph is that it shows (in a separate graph for 18 species) the
total, divided into the three main phases of breeding, being: 1)
pre-laying, 2) incubation and nestling periods and 3) fledgling and
dependent young periods. This is shown by month and number of breeding observations,
importantly divided by the number of observer weeks in the month, this
gives a true ratio, rather than just presenting raw counts. Raw counts are very
skewed by variations in the number of observer weeks over the year. This is
proved by the dip in the middle of Martin's graph, which is totally
due to the fact that many observers are away in the last week of December (as
shown in Figure 6 of the GBS Report), and nothing whatever to do with the birds'
behaviour.
As for the comment: "there is not a single CO record in the
database for this species. Clearly they appreciate their privacy!" The
conclusion, whilst presumably offered in humour, does not demonstrate an
understanding of the real reason. The GBS
Report gives the history. Which is: after my review of the GBS in 1988-90, I and
the records group of COG instituted the instruction to include the new set of
breeding codes, with Version 3 of the chart in 1993. This was to match the new
codes that the COG records system had developed and bring both into line. Before
1993 there is not likely to be any co obs for any species (because the code
was not listed as an option). More importantly, the instructions say: "Record
any breeding with the first applicable of any of the following
codes: ny; nest with young, ne; nest with eggs, fs; faecal
sac, on; bird on or seen leaving nest, nb; nest building,
cf; carrying food, dy; dependent young, ih; inspecting
hollow, di; display, co; copulation, or keep a separate
list." Note that "the first applicable" is instructed and
that co is the last given option. This sequence is not
random, it was carefully thought through, to give strongest weight to evidence
that indicates that breeding is actually occurring (preferably locally). Seeing
birds copulate, is not of itself, terribly strong evidence of local breeding, in
contrast say, to nest with young, that is definite evidence both of time
and location. Nest with young is listed before nest with eggs because it is
evidence and that it has moved one step closer to success than nest with eggs.
Likewise the reason dy is not listed first (indicating success)
is that it does not, of itself, show that the nest was within the GBS area. That
is why there is not a lot of co observations for any species (being listed
last). The space available in the little boxes on the chart is limited.
Observers should record the most definitive evidence of local breeding that is
available. It has more to do with intelligent design of the instructions on the
chart than that the birds mynas or anyone else appreciate their
privacy!
As a last comment, Martin also wrote: "I have
trawled the breeding records in the GBS database". I mention, that because
of the way I designed the GBS database, it should take about 1 minute for
someone to output that information.
Philip
|
