[BCNnet] Birds and science from Steve Bailey
Sheryl DeVore
sdevore@voyager.net
Mon, 27 Jan 2003 13:52:44 -0800
This is a multi-part message in MIME format.
------=_NextPart_000_002F_01C2C60B.5DD86D80
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
This message from Steve Bailey:
It is nice to see others talking about science as it relates to birds!
Although I would love to chime in on the latest topic, I will refrain from
doing so. However, I would like to re-visit a discussion on scientific
rigor and what conclusions can or should be drawn from any work which is
purported to be science. A recent, extremely worthwhile citizen scientist
program survey, albeit extremely limited in its scope though it was, found
reduced numbers of two bird species in the Chicago area. The surveys for
one species encompassed a grand total of 11 whole hours of censusing, while
the other was somewhat more "thorough" in its coverage at 70-80 hours. To
give you some frame of reference here, I spend ~ 8 hours conducting 2 USF&WS
Breeding Bird Surveys on parts of two mornings each summer. Even if I found
absolutely zero individuals for a number of different species on these
surveys, I would not presume that all of these species were absent from the
"search area", let alone that a single agent might be responsible for these
species demise (no matter how many folks may have died that year from car
accidents in the same area). I also monitor bird populations statewide in
my job and spend a few to several hundred hours doing so each summer. Over
the course of six years of fieldwork involving censuses conducted at around
425 different sites in 100 or 101 of the 102 counties in Illinois, my
colleague and I have found Cerulean Warblers at a grand total of two sites
involving three individual Cerulean Warblers. (Of course I failed to
mention that only a third of these sites were forest sites!) Does this mean
that Cerulean Warblers are nearly extinct in Illinois as a breeding bird? I
don't think so. Does this mean that Cerulean Warblers are in trouble in
Illinois? Probably. Does this mean that only our research alone should be
held up as THE research that gets the Cerulean Warbler placed on the Federal
List of Endangered Species? I hardly think so! With the hundreds of hours
and years of research that we have conducted, do we know the exact cause of
the decline of a species that at one time was a common component of Illinois
forests? Not by a long shot, although we do have a few good, working
theories involving forest fragmentation, Brown-headed Cowbird parasitism,
etc. which are based on thousands and thousands of research conducted by a
whole host of different researchers over a large portion of the species
range, spanning 25+ years! It is interesting that Bob mentions the 20 YEARS
of very sound research that has been done just by the one group of
researchers on the Darwin's finches. That such good research will never be
accepted by some should tell us something. However, that also does not mean
that it takes years of work to do good research. However, it does take
sound research procedures. We have some good ideas about the Ceruleans, but
the various threats facing bird populations precludes us from giving the
EXACT cause of such a widespread decline. Now, you are likely saying to
yourself, now Steve don't be so silly, right? So why is it that 11 hours of
censusing (apparently ~ one hour by eleven different folks) and 70-80 hours
of research by other folks (again, only one to a few hours each by other
folks on one day sometime over a couple week period) has not only led to the
"fact" that these two species were virtually wiped out over a large area of
one region of the state, but also that we automatically know exactly what
did it with absolutely no further reseach into the matter? Not only that,
but major news forums such as the Washington Post, MSNBC, ABC, CBS, the
Chicago Tribune, Sun-Times and WGN have picked up the story and ran with
it!! Did these news agencies bother to give a balanced scientifically based
perspective of the matter? I hardly think so. To me it is network news
sensationalism at its finest, and only goes to make me that much more
skeptical any time I pick up a newspaper. None of the stories I read even
hinted at another posible perspective. All took the basic premise that the
"research" was "fact", almost "gospel", then only went about contacting
experts in the areas of that particular field of research, or worse yet took
other sound facts (the testing of dead birds) to further support the
extremely flimsy "research" of others.
Do any of you out there see my point here? It isn't so much the end
result that I am worried about here, it is the MEANS to the ends. For any
of you that have not already figured it out by now, I am speaking of the
"research" that was conducted in the Chicago area this past summer of crows
(11 hours worth), and chickadees (70-80 hours worth). Stories and reports
continue to be published about this "research" in spite of the concerns that
I mentioned in an early IBET post. Despite discussions that I had with
folks that were intimately involved with that project, they obviously still
do not get the point AT ALL!!
I have heard or seen very little to no details about how that data was
conducted. When trying to prove a point in concerning any "scientific
theory", a thorough discussion of methods is always a necessary
prerequisite. Not only was this discussion sadly lacking in any news story
that I encountered, but it was lacking from the reserchers themselves. For
one thing, how were the study areas picked, at random or with some
predetermination of the site in mind? If the latter, this is highly biased
and the least that could have been done, especially given the extremely
small sample size (eleven hours) of the one set of surveys, would have been
to select an equal number of areas to sample outside of the predetermined
(biased) areas. Secondly, exactly how were the surveys done? Were they all
done the same way? For instance, did one person sit at a feeder and count
the no. of birds coming to a feeder, while observer number two simply
strolled down the trail at a nearby park, while observer number three sat at
a bench in a park and counted what passed by them, while number four scoured
the park, getting off the maintained trails and covering the entire area
once (or twice)? Each sampling method is going to give you a different
number of the birds of any given species in a given area, while at least
three of the above examples could easily give you a zero total for the
species that you are looking for, given the ridicuously low amount of time
spent by any one observer in that survey. Also, what kind of habitat were
the birds searched for? Urban settings at feeders, suburban parks, in
mature forests, in secong growth forest, edge areas, overgrown fields,
prairies? All of these are going to give you a different density of the
given species you are searching for, as well as a different liklihood of
finding that species in only one hour of observation! Next, what were the
weather conditions like? One thing I do know about the study is that folks
could conduct it whenever they liked. Given that folks are likely to go for
a hike when the conditions are best, at least this may have been somewhat
standardized. However, for those that follow your feeders very closely
during the winter time, you should know by now that the better the weather
is in the winter, the fewer the birds you are going to see, or at least the
amount of time spent by those birds at the feeder is usually considerably
less. Now if the weather conditions were somewhat of a mixed bag amongst
observers, as it likely was, this also affects the results of any survey.
Wind, no wind, snow or light rain (or none), high temperature, low
temperature. This is why these are some of the basic information that is
collected on Christmas Bird Counts or any census work that is done. This is
also why the best survey protocol that one can do when conducting surveys is
to repeat the surveys on at least two or three (or more) days through the
period. In this way, the affects of a weather anomoly can be filtered out
somewhat in the overall results of a census.
Now it appears that CBC data will be used to re-inforce the "facts" that
were derived from the extremely limited data that was collected this past
summer/fall. While personally I have little doubt that chickadees, crows
and a host of other species were affected to one extent or another, the one
thing that I do worry about is that science is not given a bad name, by
jumping much too quickly to conclusions in this scenario. At least in one,
if not both of the above bird species's cases, there is likely more than one
factor contributing to the reduction of these species's numbers. I could go
on for another page or two (or more!) about some likely cause and affect
scenarios for this situation. Instead, I would like to point out something
from some recent research conducted by colleagues of mine. They have been
conducting bird surveys on birds wintering in a particulat part of Illinois
in winter for each of the last four winters now. Hundreds of hours were
spent conducting very standardized census point counts, such that there is
basically no bias in the data, and as such, the results should represent
bird populations in that area at that time of year pretty accurately. To
make a long story short, when this data was compared with Christmas Bird
Count (CBC) data, there were some interesting results. Although some
species data compared pretty favorably with CBC data, for several species
for which the census data gave very stable population numbers each and every
year, numbers of those same species from CBC data, in the SAME AREAS where
the independent surveys were being conducted (and even in the SAME YEARS
when both the CBCs and censuses were conducted!), there were great
fluctuations in those species numbers in any given year, when in reality,
there wasn't such flucuation going on! This can be accounted for by the
very unstandardized ways in which CBCs are conducted, as well as the
anomolies of weather on count day, etc.. Does this mean that CBC data is
not worthwhile? Not at all. It simply means that such biases should be
realized by anyone trying to use such data to make any kind of broad,
sweeping generalizations from the data.
It will be interesting to see how research from future data plays out.
I hope that more is done in determining the true causal relationships among
West Nile Virus and the deaths of our bird species to see how best we can
help them in this matter. Otherwise, birds will continue to die long after
a vaccine is found for humans, and these deaths may or may not have anything
to do directly with West Nile Virus. When bird deaths were first being
attributed to TV towers, it was immediately felt that the towers themselves
were the cause of the deaths. Upon a little more investigation, it was
found that the height of the tower, the guide wires which supported the
towers, as well as the colors of the lights blinking on and off, had a lot
more to do with how many birds were being killed than the tower itself.
Folks need to dig a little deeper before jumping to too many conclusions,
especially when it is being held up in the name of science.
Steve Bailey
Rantoul (Champaign Co.)
sdbailey@mail.inhs.uiuc.edu
------=_NextPart_000_002F_01C2C60B.5DD86D80
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Diso-8859-1">
<META content=3D"MSHTML 6.00.2600.0" name=3DGENERATOR></HEAD>
<BODY>
<DIV><FONT face=3DArial size=3D2>
<DIV><FONT face=3DArial size=3D2><SPAN =
class=3D240232119-27012003><STRONG>This message=20
from</STRONG> <STRONG>Steve Bailey</STRONG>:</SPAN></FONT></DIV>
<DIV><FONT face=3DArial size=3D2><SPAN=20
class=3D240232119-27012003></SPAN></FONT> </DIV>
<DIV><FONT face=3DArial size=3D2><SPAN class=3D240232119-27012003>It is =
nice to see=20
others talking about science as it relates to birds! Although I =
would love=20
to chime in on the latest topic, I will refrain from doing so. =
However, I=20
would like to re-visit a discussion on scientific rigor and what =
conclusions can=20
or should be drawn from any work which is purported to be science. =
A=20
recent, extremely worthwhile citizen scientist program survey, albeit =
extremely=20
limited in its scope though it was, found reduced numbers of two bird =
species in=20
the Chicago area. The surveys for one species encompassed a grand =
total of=20
11 whole hours of censusing, while the other was somewhat more =
"thorough" in its=20
coverage at 70-80 hours. To give you some frame of reference here, =
I spend=20
~ 8 hours conducting 2 USF&WS Breeding Bird Surveys on parts of two =
mornings=20
each summer. Even if I found absolutely zero individuals for a =
number of=20
different species on these surveys, I would not presume that all of =
these=20
species were absent from the "search area", let alone that a single =
agent might=20
be responsible for these species demise (no matter how many folks may =
have died=20
that year from car accidents in the same area). I also monitor =
bird=20
populations statewide in my job and spend a few to several hundred =
hours=20
doing so each summer. Over the course of six years of fieldwork =
involving=20
censuses conducted at around 425 different sites in 100 or 101 of the =
102=20
counties in Illinois, my colleague and I have found Cerulean Warblers at =
a grand=20
total of two sites involving three individual Cerulean Warblers. =
(Of=20
course I failed to mention that only a third of these sites were forest=20
sites!) Does this mean that Cerulean Warblers are nearly extinct =
in=20
Illinois as a breeding bird? I don't think so. Does this =
mean that=20
Cerulean Warblers are in trouble in Illinois? Probably. Does =
this=20
mean that only our research alone should be held up as THE research that =
gets=20
the Cerulean Warbler placed on the Federal List of Endangered =
Species? I=20
hardly think so! With the hundreds of hours and years of research =
that we=20
have conducted, do we know the exact cause of the decline of a species =
that at=20
one time was a common component of Illinois forests? Not by a long =
shot,=20
although we do have a few good, working theories involving forest =
fragmentation,=20
Brown-headed Cowbird parasitism, etc. which are based on thousands and =
thousands=20
of research conducted by a whole host of different researchers over =
a large=20
portion of the species range, spanning 25+ years! It is =
interesting that=20
Bob mentions the <STRONG>20 YEARS</STRONG> of very sound research =
that has=20
been done just by the one group of researchers on the Darwin's =
finches. =20
That such good research will never be accepted by some should tell us=20
something. However, that also does not mean that it takes years of =
work to=20
do good research. However, it does take sound research =
procedures. We=20
have some good ideas about the Ceruleans, but the various threats facing =
bird=20
populations precludes us from giving the EXACT cause of such a =
widespread=20
decline. Now, you are likely saying to yourself, now Steve don't =
be so=20
silly, right? So why is it that 11 hours of censusing (apparently =
~ one=20
hour by eleven different folks) and 70-80 hours of research by other =
folks=20
(again, only one to a few hours each by other folks on one day sometime =
over a=20
couple week period) has not only led to the "fact" that these two =
species were=20
virtually wiped out over a large area of one region of the state, but =
also that=20
we automatically know exactly what did it with absolutely no further =
reseach=20
into the matter? Not only that, but major news forums such as the=20
Washington Post, MSNBC, ABC, CBS, the Chicago Tribune, Sun-Times and=20
WGN have picked up the story and ran with it!! Did these news =
agencies bother to give a balanced scientifically based perspective of =
the=20
matter? I hardly think so. To me it is network news =
sensationalism=20
at its finest, and only goes to make me that much more skeptical any =
time I pick=20
up a newspaper. None of the stories I read even hinted at another =
posible=20
perspective. All took the basic premise that the "research" was =
"fact",=20
almost "gospel", then only went about contacting experts in the areas of =
that=20
particular field of research, or worse yet took other sound facts (the =
testing=20
of dead birds) to further support the extremely flimsy "research" =
of=20
others.</SPAN></FONT></DIV>
<DIV><FONT face=3DArial size=3D2><SPAN =
class=3D240232119-27012003> =20
Do any of you out there see my point here? It isn't so much the =
end result=20
that I am worried about here, it is the MEANS to the ends. For any =
of you=20
that have not already figured it out by now, I am speaking of the =
"research"=20
that was conducted in the Chicago area this past summer of crows (11 =
hours=20
worth), and chickadees (70-80 hours worth). Stories and reports =
continue=20
to be published about this "research" in spite of the concerns that I =
mentioned=20
in an early IBET post. Despite discussions that I had with folks =
that were=20
intimately involved with that project, they obviously still do not get =
the point=20
AT ALL!! </SPAN></FONT></DIV>
<DIV><FONT face=3DArial size=3D2><SPAN =
class=3D240232119-27012003> I=20
have heard or seen very little to no details about how that data was=20
conducted. When trying to prove a point in concerning any =
"scientific=20
theory", a thorough discussion of methods is always a necessary=20
prerequisite. Not only was this discussion sadly lacking in any =
news story=20
that I encountered, but it was lacking from the reserchers =
themselves. For=20
one thing, how were the study areas picked, at random or with some=20
predetermination of the site in mind? If the latter, this is =
highly biased=20
and the least that could have been done, especially given the extremely =
small=20
sample size (eleven hours) of the one set of surveys, would have been to =
select=20
an equal number of areas to sample outside of the predetermined (biased) =
areas. Secondly, exactly how were the surveys done? Were =
they all=20
done the same way? For instance, did one person sit at a feeder =
and count=20
the no. of birds coming to a feeder, while observer number two simply =
strolled=20
down the trail at a nearby park, while observer number three sat at a =
bench in a=20
park and counted what passed by them, while number four scoured the =
park,=20
getting off the maintained trails and covering the entire area once (or=20
twice)? Each sampling method is going to give you a different =
number of=20
the birds of any given species in a given area, while at least =
three of the=20
above examples could easily give you a zero total for the species that =
you are=20
looking for, given the ridicuously low amount of time spent by any one =
observer=20
in that survey. Also, what kind of habitat were the birds searched =
for? Urban settings at feeders, suburban parks, in mature forests, =
in=20
secong growth forest, edge areas, overgrown fields, prairies? All =
of these=20
are going to give you a different density of the given species you are =
searching=20
for, as well as a different liklihood of finding that species in only =
one hour=20
of observation! Next, what were the weather conditions like? =
One=20
thing I do know about the study is that folks could conduct it whenever =
they=20
liked. Given that folks are likely to go for a hike when the =
conditions=20
are best, at least this may have been somewhat =
standardized. =20
However, for those that follow your feeders very closely during the =
winter time,=20
you should know by now that the better the weather is in the winter, the =
fewer=20
the birds you are going to see, or at least the amount of time spent by =
those=20
birds at the feeder is usually considerably less. Now if the =
weather=20
conditions were somewhat of a mixed bag amongst observers, as it likely =
was,=20
this also affects the results of any survey. Wind, no wind, snow =
or light=20
rain (or none), high temperature, low temperature. This is why =
these are=20
some of the basic information that is collected on Christmas Bird Counts =
or any=20
census work that is done. This is also why the best survey =
protocol that=20
one can do when conducting surveys is to repeat the surveys on at least =
two or=20
three (or more) days through the period. In this way, the affects =
of a=20
weather anomoly can be filtered out somewhat in the overall results of a =
census.</SPAN></FONT></DIV>
<DIV><FONT face=3DArial size=3D2><SPAN =
class=3D240232119-27012003> =20
Now it appears that CBC data will be used to re-inforce the "facts" that =
were=20
derived from the extremely limited data that was collected this past=20
summer/fall. While personally I have little doubt that chickadees, =
crows=20
and a host of other species were affected to one extent or another, the =
one=20
thing that I do worry about is that science is not given a bad name, by =
jumping=20
much too quickly to conclusions in this scenario. At least in one, =
if not=20
both of the above bird species's cases, there is likely more than one =
factor=20
contributing to the reduction of these species's numbers. I could =
go on=20
for another page or two (or more!) about some likely cause and =
affect=20
scenarios for this situation. Instead, I would like to point out =
something=20
from some recent research conducted by colleagues of mine. They =
have been=20
conducting bird surveys on birds wintering in a particulat part of =
Illinois in=20
winter for each of the last four winters now. Hundreds of hours =
were spent=20
conducting very standardized census point counts, such that there is =
basically=20
no bias in the data, and as such, the results should represent bird =
populations=20
in that area at that time of year pretty accurately. To make =
a long=20
story short, when this data was compared with Christmas Bird Count (CBC) =
data,=20
there were some interesting results. Although some species data =
compared=20
pretty favorably with CBC data, for several species for which the census =
data=20
gave very stable population numbers each and every year, numbers of =
those same=20
species from CBC data, in the SAME AREAS where the independent surveys =
were=20
being conducted (and even in the SAME YEARS when both the CBCs and =
censuses were=20
conducted!), there were great fluctuations in those species numbers in =
any given=20
year, when in reality, there wasn't such flucuation going on! This =
can be=20
accounted for by the very unstandardized ways in which CBCs are =
conducted, as=20
well as the anomolies of weather on count day, etc.. Does this =
mean that=20
CBC data is not worthwhile? Not at all. It simply means that =
such=20
biases should be realized by anyone trying to use such data to make any =
kind of=20
broad, sweeping generalizations from the data.</SPAN></FONT></DIV>
<DIV><FONT face=3DArial size=3D2><SPAN =
class=3D240232119-27012003> =20
It will be interesting to see how research from future data plays =
out. I=20
hope that more is done in determining the true causal relationships =
among West=20
Nile Virus and the deaths of our bird species to see how best we can =
help them=20
in this matter. Otherwise, birds will continue to die long after a =
vaccine=20
is found for humans, and these deaths may or may not have anything to do =
directly with West Nile Virus. When bird deaths were first being=20
attributed to TV towers, it was immediately felt that the towers =
themselves were=20
the cause of the deaths. Upon a little more investigation, it was =
found=20
that the height of the tower, the guide wires which supported the=20
towers, as well as the colors of the lights blinking on and off, =
had a lot=20
more to do with how many birds were being killed than the tower =
itself. =20
Folks need to dig a little deeper before jumping to too many =
conclusions,=20
especially when it is being held up in the name of science. =20
</SPAN></FONT></DIV>
<DIV><FONT face=3DArial size=3D2><SPAN =
class=3D240232119-27012003></SPAN></FONT><FONT=20
face=3DArial size=3D2><SPAN =
class=3D240232119-27012003></SPAN></FONT> </DIV>
<DIV><FONT face=3DArial size=3D2><SPAN class=3D240232119-27012003>Steve=20
Bailey</SPAN></FONT></DIV>
<DIV><FONT face=3DArial size=3D2><SPAN =
class=3D240232119-27012003>Rantoul (Champaign=20
Co.)</SPAN></FONT></DIV>
<DIV><FONT face=3DArial size=3D2><SPAN class=3D240232119-27012003><A=20
href=3D"mailto:sdbailey@mail.inhs.uiuc.edu">sdbailey@mail.inhs.uiuc.edu</=
A></SPAN></FONT></DIV></FONT></DIV></BODY></HTML>
------=_NextPart_000_002F_01C2C60B.5DD86D80--