[Leps-l] moth sampling

Tue Jan 20 12:57:36 EST 2015

I think that this also has to do with how such indicators are misused (myself included).  If I recall, Chao2 was designed to tell you how many species of zooplankton/phytoplankton were present in a body of water on a given sampling event.   So - a finite species pool, in one habitat on one day.  Compare this to what I'm doing, sampling across years (high species turnover by season) and habitat types in a small-ish country - and you have a mess.  But what else do we have?

I think these types of estimators might work fairly well for moth sampling where traps are located in the same place every sample event, and a year's worth of sampling is "a sample" (analogous to a set of plankton net pulls from a lake).

j

  Please consider the environment before printing this email 

John A Shuey
Director of Conservation Science

jshuey at tnc.org
317.829.3898 - direct
317.951.8818 - front desk 
317.917.2478 - Fax 

nature.org 

The Nature Conservancy
Indiana Field Office 
620 E. Ohio St.
Indianapolis, IN 46202

-----Original Message-----
From: Walsh, James Bruce - (jbwalsh) [mailto:jbwalsh at email.arizona.edu] 
Sent: Tuesday, January 20, 2015 12:32 PM
To: John Shuey; Tony Thomas; leps-l at mailman.yale.edu
Subject: RE: [Leps-l] moth sampling

"Perhaps this is a solid indication that we are no-where close to having a decent inventory?"
Bingo!
________________________________________
From: John Shuey [jshuey at TNC.ORG]
Sent: Tuesday, January 20, 2015 9:12 AM
To: Walsh, James Bruce - (jbwalsh); Tony Thomas; leps-l at mailman.yale.edu
Subject: RE: [Leps-l] moth sampling

Hi Bruce,  In some way, this is like the Chao estimators.  The problem with these (at least relative to this particular data base) is that no matter how you cut them (sampling effort-wise), the estimates always keep going up.   Every new field effort knocks 20 or so species of the "single record list" but adds 20 or so new species to that list.  In the ideal world, at some point, the number of unique records has to start leveling off - which for Belize butterflies - we can't get it to do that.  In other words, for every unique collection record species we knock off the list, we replace it with a new unique single record-species.   I want an indicator that points to a fairly consistent "predicted total # of species"    - a stable predicted asymptote that I can approach as sampling effort increases.  Perhaps this is a solid indication that we are no-where close to having a decent inventory?

It's frustrating, because I want Belize to be over, so that we can move onto something new and exciting.

j

  Please consider the environment before printing this email

John A Shuey
Director of Conservation Science

jshuey at tnc.org
317.829.3898 - direct
317.951.8818 - front desk
317.917.2478 - Fax

nature.org

The Nature Conservancy
Indiana Field Office
620 E. Ohio St.
Indianapolis, IN 46202

-----Original Message-----
From: Walsh, James Bruce - (jbwalsh) [mailto:jbwalsh at email.arizona.edu]
Sent: Tuesday, January 20, 2015 9:52 AM
To: John Shuey; Tony Thomas; leps-l at mailman.yale.edu
Subject: RE: [Leps-l] moth sampling

By the way, a quick "adjustment" (a delete-one jackknife estimator) for the total number of species is total number seem PLUS the number of species seen in only one sampling event (not singletons, but rather seen on only one day, could be hundreds that day).  For this data, add 145 and that gives you a better estimate of the actual number of species

cheers

bruce
________________________________________
From: leps-l-bounces at mailman.yale.edu [leps-l-bounces at mailman.yale.edu] on behalf of John Shuey [jshuey at TNC.ORG]
Sent: Tuesday, January 20, 2015 6:35 AM
To: Tony Thomas; leps-l at mailman.yale.edu
Subject: Re: [Leps-l] moth sampling

Hi Tony,  I think your data are interesting and reflect a somewhat "mature" survey end point. - you are at the 39 records/species point.  Here is what I have at the 29 records/species mark for Belize butterflies at the (27,954 record for 958 species - hence a 25% less mature survey in my mind).

406 species known from 5 or fewer records (note that in this accounting, a record may a more than one specimen - it's a measure of a collection of a species at a place on a unique date, irrespective of how many individuals were recorded):

145 species known from single capture events
91 species known from 2 events:
61 species known from 3 records:
44 from 4 records;
65 from 5 records
One species has 459 records

The similarities are striking - no?

Our data set is probably about 4,000 records behind, and the change in the shape of the graph should be interesting once we decide to pull the plug and claim that we are "done".  But I hope to see the number of rarely encountered species decline as a function of "time" (as measured by # of records)

j

  Please consider the environment before printing this email

John A Shuey
Director of Conservation Science

jshuey at tnc.org
317.829.3898 - direct
317.951.8818 - front desk
317.917.2478 - Fax

nature.org

The Nature Conservancy
Indiana Field Office
620 E. Ohio St.
Indianapolis, IN 46202

-----Original Message-----
From: leps-l-bounces at mailman.yale.edu [mailto:leps-l-bounces at mailman.yale.edu] On Behalf Of Tony Thomas
Sent: Monday, January 19, 2015 4:46 PM
To: leps-l at mailman.yale.edu
Subject: [Leps-l] moth sampling

Hi All,

Tidying up my computer's hard drive I came across a graph from my  moth diversity work in Fundy National Park in New Brunswick, Canada; many years ago.
Out of 401 species of macros with 15,851 specimens, about half (approx. 190 species) had 5 or less individual moths The number of species with 1, 2, 3, 4, 5 moths was 75, 40, 35, 20, 20 respectively. At the other extreme one species had 1,253 individuals.
I believe this demonstrates that for the best inventory of species one should collect every night if one wishes to collect the very rare species (i.e., those with only 1 individual that comes to a trap).
Of course my data doesn't rule out the possibility that if  traps are operated just once every 5 days the singleton that would have been caught on day 2, when traps are operated nightly,  may be caught on day 5.

_______________________________________________
Leps-l mailing list
Leps-l at mailman.yale.edu
http://mailman.yale.edu/mailman/listinfo/leps-l
_______________________________________________
Leps-l mailing list
Leps-l at mailman.yale.edu
http://mailman.yale.edu/mailman/listinfo/leps-l