[CT Birds] Varied Thrush, final words
dennisvz at optonline.net
Fri Jan 8 19:07:06 EST 2010
The first step in analyzing any data is knowing your data set. How it
is collected and what is available determines what you can do with
it. Initially all I was interested in was: Was the impression of a
decline in Varied Thrush records real or an illusion. With the
information I had at hand it appeared to be true, but I was unsure of
the recent records. With the inclusion of new records, the impression
was indeed an illusion. The cause of the illusion was the lack of
general reporting of birds in residential areas in recent years.
Dave Provencher presented many ideas about vagrancy though true
were not relevent in this case.
To do this "study" I grouped the data into 10 year intervals to
remove inter-year variation. By clustering, many sources of variation
get averaged out such as weather patterns, breeding success, food
availability etc. If one of these variables was consistent over the
past fifty years then they would affect population size and show up
in the pattern.
Another source of variation is randomness. The implicit assumption is
made that birds are sampled at random. So, Mass. NY, RI, are as
equally likely to get birds as Connecticut and is proportional to
land area. Another assumption is that when random sampling, the
number of birds found is proportional to the number of birds present
(not the actual number present). These assumptions may be false but
it would take another type of analysis to test them (all records seem
to be from the Conn. R. West) Out of curiosity I took a look at
Record of New England Birds 1946 to 1953 and there are no reports.
John Bull Birds of NYC lists 3 records from before 1900 and the
following 1905, 1928, 1936, 1958, 1960, and 1963.
As it now stands in Connecticut
1960 to 1969 3 records
1970 to 1979 4 records
1980 to 1989 5 records
1990 to 1999 6 records
2000 to 2009 5 records
2006/3 to 4 New Hartford
2008 /9 to 12 Wilton (available to public)
More information about the CTBirds