- As an ecologist interested in the role of flowering plant diversity and provenance on insect biodiversity and biocontrol, I ve found this discussion important,Message 1 of 8 , Sep 8, 2010View Source
As an ecologist interested in the role of flowering plant diversity and provenance on insect biodiversity and biocontrol, I’ve found this discussion important, provocative, and instructive. As natural and semi-natural habitats are assaulted from all sides by development and pollution, we must have defensible data on pollinator and other insect biodiversity to inform scientific and public discussion of these threats. Instruments for the necessary surveys seem to run from nets to malaise traps: the former is biased by investigator behavior; the latter seems a blunt-edged battering ram with unknown biases, including, importantly, the effective capture radius.
I’m new to this area but my impression from discussions with various hymenopterists is that pan trapping offers the best combination of thoroughness and spatial circumscription for tying species richness to particular habitats. To the extent possible one should limit trapping to the lowest duration and frequency for the task (perhaps this is known for at least some groups from cumulative diversity curves). One must also be mindful of the by-catch, which includes many other groups – e.g., parasitic and predatory wasps, predatory beetles – that are also important to ecosystem function. In the case of the non-pollinating Hymenoptera, many systematists would be eager for a chance to have at them; in general, we can all save our by-caught material in EtOH in cool, dark places until we can find the right person to give them to.
From: email@example.com [mailto:firstname.lastname@example.org] On Behalf Of Strange, James
Sent: Tuesday, September 07, 2010 7:56 PM
To: Sam Droege
Subject: RE: [beemonitoring] On squash bee, surveys, and the death of insects
Sam et al.,
Thanks for your thoughtful posts. It is indeed a diverse group that works on bees.
I hate to pick at scabs, but during the flurry of posts last week I thought it best to walk away and let the holiday weekend temper my thoughts on this issue. I do want to throw my two cents into the pot as I think this is an important issue. This is, of course, not about Peponapis or even traps, really. It is about bees and the ethics of killing them. And, as it turns out, the easiest way to kill and collect oodles of bees is to trap them. As Sam rightfully pointed out, we are all going to approach this in different ways, with different feelings about this and, I suspect, those feelings will largely be influenced by the particular work the individual is doing. And, yes, it depends on our objectives. It is easy for me to justify the taking of several hundred bees for pathogen or population surveys because I “need” to kill the bees to acquire the data. I imagine we all perceive similar “needs” for our own work.
I don’t suppose we need some ethical treatise on the subject of killing bees, but I think that we should be able to bring up the question without too much rancor. A few simple checks (What are my research goals? How many specimens do I need to take to achieve those goals? What are the ups and downs of my particular methodology?) should be employed before any project. I would add a few other questions to the list. If I kill bees, how can I maximize the data they generate? Could I justify this to an ethical review board? What would Vince do?
I personally don’t find it necessary to know all of the bees that exist in the world; my interests seem to revolve around the more common ones. However, if you are so inclined, then I have no major beef with your goals. Yet, I have lately seen several collections that have come out of traps (of various kinds) and I am in general disappointed that the collectors used that method. For large faunal surveys, they may be an excellent tool, but too many people (and sadly many are students) are using the traps to gather data where trapping is inappropriate. As Peter mentioned, the pollen is often gone, sometimes the specimens are poorly treated after recovery, or in the case of the vane traps the specimens are broken and brittle. There is also the issue of by-catch to contend with. All those unwanted beetles, moths and flies add up. If the particular trap (e.g. Alex’s Peponapis trap) can reduce the by-catch then I am in favor. Though I would still argue that in most cases the bees could be observed at the flowers and captured if you are really interested in the visitors to a particular flower. Perhaps there are other cases where it would be even better applied.
In general I take the philosophy that Jim Cane and Vince Tepedino espoused, which is to say I have reservations about large scale faunal surveys in general. Where a clear scientific question is posed a priori and the benefits can be seen to outweigh the costs, then by all means, go for it. While it is interesting to consider the possibility of what we might be able to do if we had a few long series of Bombus affinis as John pointed out, I still have a hard time wishing there were more in collections. The best series we have of Bombus occidentalis came from Macior’s net collections while studying the pollination of Pedicularis, not from large faunal surveys. Unfortunately his labeling leaves something to be desired. A final thought to chew on: What if a trap turns up a dead Bombus franklini queen? An undergrad with a net could be trained to photograph and release that same queen. Would trapping be justified to say “look, the species was extant last week”?
Thanks for your time and your thoughts,
James P. Strange PhD
What interesting and passionate people we have on this listserv. This current thread reminds me of the differences between two of my favorite poets the Russian American amateur butterfly taxonomist Vladimir Nabokov and the nature observant Emily Dickinson. Their philosophies differ in similar ways (poems at end of this email.).
There are two components being discussed here, ..The death of bees via our collecting actions and the best approach to a survey of bees in a research, inventory, or monitoring situation.
How we approach the moral issue regarding the death of insects is a personal one, we all have differences in what we think, how we feel, and how we transmit those feelings about our direct actions that lead to an insect's death. Philosophically and logically, there is no right or wrong here, there are simply differences, we can argue and persuade, but in the end no statement can be made outside our personal feelings. Because it involves passions and feelings its best to be careful in how it is discussed and is similar to the arenas of religion and politics in what it can do to relationships.
In terms of the use of a trapping technique (for example the flower trap that stimulated this discussion) there are some statistical and mathematical guidelines that can be helpful in this discussion. In fact, I have just returned from teaching a week long course (with several FWS statisticians) at the National Conservation Training Center on developing surveys for animals and plants and we worked with everything from Caribou to Spring Snails. So these ideas are fresh in my mind.
It all depends on your objectives.
In the broadest sense a good survey methodology demonstrates low methodological variance and low bias. In other words you would like a strong relationship between the number of things counted or captured and the actual number of things out there you want to talk about and that your technique of choice does that efficiently and is not adding a huge amount of noise to those counts (which would require boosting sample sizes to achieve the same precision). This relationship is never perfect and is almost never evaluated. For example, off the top of my head I can't recall any survey of bees using any technique that evaluated whether their visual counts, nettings, trappings were unbiased in relationship to the real populations as it is very difficult to either estimate or determine how many bees are really in a location at any moment in time (and lots of additional difficulties of even defining things as simple as "area" or "density," but I am off topic).
At the next level, your objectives drive what survey technique is used and how bees are sampled and consequently, how many bees are sacrificed. So, in the case of the flower trap it may be the least variable and most unbiased way to survey bees for any given crop or situation or it may not be. We do not know as no tests have been made nor has the technique been extensively deployed and observed to see if Alex's results are replicable and what factors might contribute to how many bees go into the trap and what the sex rations and species compositions become. We can't know until those surveys are done. At this point Alex (in my mind) has made a welcome and major contribution in discovering this phenomenon and now we must test it to see how well it works out well statistically in other areas and perhaps has applications for difficult to survey crops like Apple. It makes sense that flower traps may be very useful to the study of flowering plants as they would appear to better target the bees of interest, increase capture rates in situations where captures are low, permit accurate and verifiable species identification, have low observer bias, and are inexpensive. On the negative side they may (or may not....) kill many bees within a study area and that may conflict with goals for retaining bees on site or simply be in conflict with the moral principles of the project and participants.
Again at a broad level there are some generalities that can be applied to survey techniques.
Traps are less affected by observer bias than counts and net captures and are usually cheaper and easier to obtain large sample sizes per unit effort.
All else being equal large numbers of counts or captures are mathematically more useful in a monitoring or research study than low numbers as when you have low numbers your sampling distribution runs in the zero wall as you cannot have negative animals and the calculations of variances and sample noise is much larger when dealing with zeros. The difference between 0 and 1 is much greater than 10 and 11. With low counts you can your ability to detect difference and are required to collect more samples to make up those differences or pool across samples which ultimately may mean increasing your effort and even the number of insects collected to overcome there statistical issues.
On Discovering a Butterfly
I found it and I named it, being versed
in taxonomic Latin; thus became
godfather to an insect and its first
describer -- and I want no other fame.
Wide open on its pin (though fast asleep),
and safe from creeping relatives and rust,
in the secluded stronghold where we keep
type specimens it will transcend its dust.
Dark pictures, thrones, the stones that pilgrims kiss,
poems that take a thousand years to die
but ape the immortality of this
red label on a little butterfly.
-- Vladimir Nabokov
ARCTURUS is his other name,—
I ’d rather call him star!
It ’s so unkind of science
To go and interfere!
I pull a flower from the woods,— 5
A monster with a glass
Computes the stamens in a breath,
And has her in a class.
Whereas I took the butterfly
Aforetime in my hat, 10
He sits erect in cabinets,
The clover-bells forgot.
What once was heaven, is zenith now.
Where I proposed to go
When time’s brief masquerade was done, 15
Is mapped, and charted too!
What if the poles should frisk about
And stand upon their heads!
I hope I ’m ready for the worst,
Whatever prank betides! 20
Perhaps the kingdom of Heaven ’s changed!
I hope the children there
Won’t be new-fashioned when I come,
And laugh at me, and stare!
I hope the father in the skies 25
Will lift his little girl,—
Old-fashioned, naughty, everything,—
Over the stile of pearl!
- Emily Dickinson