Fwd: Re: buttcounts & other schmetterkopfery

[Chris J. Durden]

>Date: Wed, 01 May 2002 11:32:59 -0500
>To: TILS-leps-talk at yahoogroups.com
>From: "Chris J. Durden" <drdn at mail.utexas.edu>
>Subject: Re: buttcounts & other schmetterkopfery
>
>For many years (seriously since 1960) I have attempted to de-bias my 
>collecting data by using a "first up" protocol. Catch, or count the first 
>individual seen when looking up from dealing with the last one. The 
>resulting numbers are certainly not unbiased - one chooses where to sample 
>and how hard to work, but I do think the numbers can be used for 
>inter-locality and inter-year comparison.
>    At the 1981 LepSoc/SocMexLep meeting at Cuautla, Bob Pyle persuaded me 
> to start a Xerces count at Austin TX. In 1982 I did and we had excellent 
> participation in the years before 1986 when the invading Fire Ants 
> (*Solenopsis invicta*) decimated our fauna. In years since our counts 
> have indicated a slow adjustment to living with the Fire Ants, although 
> our numbers are half what we formerly had, our diversity is half what we 
> formerly had, and our commonest species are now scarce. Every year 
> (except one missed) the same original Xerces guidelines are followed so 
> the numbers can be directly compared. When NABA hijacked the Xerces 
> counts, this count went private after discovering the amount of needless 
> busywork needed to translate Xerces-style data into NABA format, with the 
> associated loss of precision. Our Austin count is conducted the first 
> Saturday during the interval day 190 to 199. This is our midsummer which 
> in drought years is quite a bottleneck and during mild summers can 
> produce high diversity. Although absolute numbers are much below what we 
> could find in Spring and Fall. Numbers in Fall would be overwhelming and 
> I would not trust any sight count taken at that time. It is just 
> physically impossible to count or estimate all butterflies in the field 
> of view at one moment in a meadow of *Viguiera* or a hillside of 
> *Eupatorium havanense*. Numerical data taken at these times of abundance 
> are by sampling, using "first up" protocol with a slight bias towards 
> rarity when an hurricane waif enters the field of view.
>    In working with the numerical data I find it useful to normalize data 
> for each species, for inter-year comparison. Each species is scored for 
> its portion of the total count of all species of HLPNP, the five 
> butterfly groups (diurnal moths are counted but not included in the 
> butterfly total). From these frequency numbers is determined the rank 
> number for each species. These rank numbers are compared from year to 
> year. It seems to me that the rank number will be a fairly accurate 
> indicator of relative species success from year to year whatever the 
> weather pattern and whatever the skill and participation of the counters.
>    It is interesting to compare sites at different elevations and 
> latitude, using this normalized data. Plot the log of rank number against 
> the log of portion of the whole represented by each species. The result 
> is a graphic ranking of species - a representation of the MacArthur 
> broken stick distribution. (Because in non-wilderness areas there is a 
> sharp bend in the curve, I refer to this as the "broken log" 
> distribution). At high latitudes and altitudes, and in early spring and 
> late fall the curve is steep. In low latitudes and mid-summer the curve 
> is shallow. In areas with much human disruption of the landscape the 
> common species are less common than expected and the scarce species are 
> rarer than expected, yielding a bend in the curve, often sharp which in a 
> fauna 0f 180 species falls at about species 15th in abundance. The high 
> latitude wilderness curve with commonest common species and scarcest 
> rarities comes closest to the Fibonacci distribution (the Bonzai growth 
> pattern, or the observed distribution of the elements in the universe). 
> The tropical wilderness curve with rarest common species and commonest 
> rare species comes closest to the "taxonomic curve" exemplified by the 
> number of genera per family in Trilobites or Ostracodes as indexed in the 
> Treatise on Invertebrate Paleontology.
>    I think butterfly count numbers can mean something if uniformly taken 
> and may ultimately be useful in calibrating the ecological "health" of 
> natural and artificial communities.
>..............Chris Durden
>
>At 12:33 PM 5/1/2002 +0000, you wrote:
>>________________________________________________________________________
>>
>>Message: 8
>>    Date: Tue, 30 Apr 2002 10:31:38 -0600
>>    From: "Barb Beck" <barb at birdnut.obtuse.com>
>>Subject: FW: Butt counts and the numbers game.
>>
>>
>>Ron and others,
>>
>>There are many ways people can survey butterflies.  Ideally we would have
>>lots of randomly placed transects manned by volunteers and surveyed
>>repeatedly over the summer.  They must be in good habitat where people can
>>find lots of butts and have fun and also in habitat where things are pretty
>>dull.
>>
>>Problem is that we are dealing with volunteers.  While a bunch will be
>>willing to run things like the random transects enough will not.  We must
>>keep the interest of volunteers. Although Alberta has (I think) the best
>>participation in North America for our population in Breeding Bird Survey
>>routes (random transects but they have to be along roads or at least
>>something we can pretend is a road) the number of participants we have in
>>these things is only  a small fraction of those who participate in the bird
>>counts and it is only necessary to run these once during the summer.  And
>>with breeding birds you can cover a LOT more territory fast (a 25 mile
>>transect with points every half mile) than with butterflies. Realize a lot
>>of the bird id particularly in forested areas is done by vocalizations and
>>you are able to estimate bird numbers from a far bigger area around each
>>sample point in a 3 minute period than you would be able to do with
>>butterflies.
>>
>>AAAAAAAAGH - just started SNOWING AGAIN!!!!!!  Are we EVER going to get
>>spring.  It is already about 4 weeks late and our warm weather is short
>>enough  as it is.  Whimper whimper...  Back to message
>>
>>For surveying a small area Pollard walks are great for butterfly study in
>>small areas, however again these things are not assigned randomly, it would
>>be difficult to get somebody to walk these things repeatedly in really
>>crummy butterfly habitat to get any kind of a rough picture about what is
>>going on province wide.
>>
>>So where do the counts come in  - admittedly without technically the best
>>experimental design but a pretty good one under the circumstances. Those
>>that are not run as small nature walks through an area of nice habitat have
>>people going out and trying to find as many species as they can.  That
>>causes them to get into some pretty miserable habitat at times or habitat
>>like our bogs which can be difficult to work in or at least to look some at
>>some pretty boring areas.  With some encouragement you can get them to
>>notice "junk" butterflies and birds as well as those people like to find.
>>That is why we keep track of "world records" we set.  It at least in the
>>CBCs it gets people giving us estimates of the Black-billed Magpies, Rock
>>Doves (Pigeons), Black-capped Chickadees and House Sparrows they find.  The
>>meaningless extra digits in the results are kept in the "records".  The
>>extra digits are kept - not for scientific accuracy but to keep people
>>reporting what they see.  Remember if we round every thing to the nearest
>>100 then some groups that only see 15 - 25 of the birds will not report.
>>Enough of these groups (we have probably about 50 parties out) that do not
>>report these things because their results will be rounded off do not realize
>>that when combined with the other small numbers for other groups they add up
>>to a significant digit.  As a scientists I would never report my own data as
>>943 Arctic Blues seen BUT when working with volunteers you add the 900
>>estimate to the 12 from another group and the 5 from another etc etc.
>>  Look - people want to see their results count.  Anybody working with the
>>data who would believe three significant figure accuracy is completely out
>>of touch with reality.  If at a tally or in the report I properly round off
>>that number people will not report the smaller numbers the next year.
>>People working with the data realize the inherent errors.   The errors are
>>far greater than a digit on a large number.  Each type of habitat is not
>>given equal coverage.  Observers vary in their abilities to detect birds.
>>The common number passed around in CBC data is a factor of 10.  BUT THAT
>>DOES NOT make the stuff worthless.  At least until pigs can fly and I am
>>covorting on the beach in a size 4 bikini we will not have the government
>>funds to hire the army of people necessary to conduct proper surveys.  This
>>gives us a picture which even though fuzzy has shown to provide some
>>valuable distribution and population data.
>>
>>Edmonton has data in the Birdsource database which goes back almost 100
>>years - our first count was in 1908.  Wouldn't it be nice to have such
>>information about our butterflies?  With a few key strokes I can get a
>>listing of all the counts a particular bird was found in  - range maps or
>>the results of a particular count over the year and all historical count
>>records.
>>
>>The counts are not randomly assigned but we are really trying to get them
>>spread out in the province.  The circles are large enough that they usually
>>have quite a sampling of the habitat available in the area.  People often
>>consciously try to get counts set up in various habitats.
>>
>>There is an element of sport in them particularly with our birdcounts .
>>There is hot competition between various zones in our Edmonton Bird count
>>circle (of course Zone 1 - the one we run is the best).  There is
>>competition between cities for whose count can turn up the most birds. We
>>frequently have challenges out with others on who can turn up the most of a
>>particular type of owl... how the heqq else do you convince sane or at least
>>partially sane people to go at midnight and spend the next 8 hours hooting
>>for owls at -25 below temperatures even if they are suffering from a severe
>>case of cabin fever.  All the challenges and bragging BS (which I am as much
>>a part of as anybody)  puts some fun and sport into what at times can be
>>some pretty cold work but we tend to  take the collection of the data pretty
>>seriously.  Basically it gets people out digging. We have found out a lot of
>>things with our CBCs. For example, since we have started to look for them
>>about 15 years ago (when some dumb Californians who did not know that
>>Saw-whet owls were not supposed to be here in the winter started finding
>>them) we have established that Northern Saw-whet owls actually overwinter
>>here in the northern part of their range in great numbers.  We have
>>established that Long-eared owls do the same as well as a lot of information
>>on our other birds. To keep people interested in reporting at least somewhat
>>reasonable estimates of common birds we keep track of yearly records and at
>>our wrapup after the day of counting we cheer as new records are set.
>>
>>  Are the people honest in what they report.  I think so.  One year we found
>>out at the wrap-up pot luck that we had tied the world record for the number
>>of Northern Saw-whet owls found.  We still have a few hours to go until
>>midnight.  The owling parties streamed out of the room back into the cold to
>>find at least one more owl.  Amazingly nobody was able to report another
>>owl. A person could have reported another owl, been hailed as a hero and
>>nobody would have been the wiser.  That pretty much convinced me that the
>>people involved were honest.
>>
>>A little bit of the sport aspect comes in with the butterfly counts. Our
>>compilers are pretty proud of the number of counts we hold (although do not
>>know if we can hold off Texas for another year)  The Alberta counters are
>>always happy to hear when they have set new records for a species in a
>>count. (unless it is an absurdity like the first ever record for a Christina
>>Sulphur in the NABA counts as they claimed in their reports this year.  We
>>have repeatedly properly turned this species under their lumped name with
>>the proper "subspecies" attached.  However that was not taken into account
>>because the Christina was just "created by God last year"  so Hinton set the
>>record and has the first ever occurrence reported - another example of lost
>>data using "conservative taxonomy" - How the heqq is this conservative?
>>(Must get off soapbox - must get off soapbox - must get off soapbox)
>>
>>Some of our butterfly counts that have the personnel are doing the same -
>>trying to get people into all possible habitats in the circle to dig out all
>>the species they can.  The high number of butterfly species picked up at our
>>Dinosaur Provincial Park count the last two years is no accident but the
>>result of some good planning by a compiler (one of my students..brag brag)
>>
>>What will you get out of these things in the long run.. presence and absence
>>of the butterflies of the region.  Long term some estimate of population
>>trends when there is enough data to obscure yearly fluctuations.  Our
>>butterfly counts have provided records of many of the species our government
>>is trying to track.  They have found the second and third record of a Summer
>>Azure in the province.  They have been tracking the spread of the introduced
>>European Skipper.  They have documented range extensions for a number of
>>species and will also in the future probably document range decreases for
>>some.  The actual numbers are crude but do give some estimate of abundance
>>per party hour.  Not perfect by a long shot but something that we can do.  A
>>fuzzy picture of what we have is better than no picture at all.  And they
>>get people out swing nets and learning about butterflies. And when requested
>>scientific specimens are collected on our counts and sent to those
>>requesting them to find out more about these creatures. At least that is
>>what goes on in Alberta.  (Please no hate mail from New England NABA types -
>>I have already heard all your chants and realize that you have problems
>>differentiating nets from shotguns. Someday I will post some field marks to
>>help you with that id problem.)
>>
>>Note for those of you not familiar with such data you heard me mention in
>>the last post the term party hour.  That is how we are able to compare
>>counts with varying amounts of participation year after year.  If we had 10
>>parties one year working for a total of 6 hours each (60 party hours) and
>>one poor guy working for 8 hours another year the data can be compared by
>>looking at the number of butterflies per party hour... no wise cracks about
>>partial butterflies please.
>>
>>The data for the CBCs is more highly filtered now than that for the
>>Butterfly counts.  NABA does not let its regional editors require
>>documentation or challenge sightings which are clearly out of wack - but I
>>always check with the compiler to make sure it is not a mistake on the
>>form - which it usually is or to hear the justification for a sighting
>>anyway.  The CBCs require rare bird forms for any species found new in a
>>count and those which are difficult to separate or very rare.  Basically
>>these tell anybody interested how the species was identified and how it was
>>differentiated from species with which it might be confused.  Of course
>>recordings and photos also often accompany these things.  The results of the
>>CBCs are published with an indication of the reliability of some of the
>>sightings.
>>
>>ANYWAY enough for this morning... looks like the snow stopped at least for a
>>while - still a little below zero.  I want to get some tea and read the
>>paper
>>
>>Barb Beck
>>Edmonton, Alberta, Canada



 
 ------------------------------------------------------------ 

   For subscription and related information about LEPS-L visit:

   http://www.peabody.yale.edu/other/lepsl