Loading ...
Sorry, an error occurred while loading the content.

2896Re: [beemonitoring] Identifying pollen collected from bees

Expand Messages
  • Berry Brosi
    Aug 23, 2013

      I agree, it should work quite well for pollinia, and yes, if following the Wilson et al. protocol each pollinium should be analyzed separately.  The issue that we don't have good coverage of plants in terms of representation in GenBank remains, and is likely particularly acute for tropical orchids.


      Berry J. Brosi, Ph.D.
      Assistant Professor
      Department of Environmental Studies
      Emory University

      On Aug 23, 2013, at 10:39 AM, Peter Bernhardt <bernhap2@...> wrote:

      Dear Berry:

      Hmmmm.  I'll bet it would work very nicely for pollinia carried by bees, though.  Each pollinium consists exclusively of pollen from one flower.  Some orchid pollinaria of the Neotropics can be identified to genus on the basis of morphology and where it is deposited on the bee's body.  The analyst knows he/she is removing pollen of Stanhopea but may not know which Stanhopea spp.  You technique will answer that question.

      Likewise, this might also be applied to milkweeds (Asclepias) and their allies.  They also produce pollinaria and, based many past studies, a bee may visit more than one Asclepias spp. in bloom during the same foraging bout.  Pollinaria of milkweeds are hard to identify once they leave the flower, and different Asclepias app. deposit the pollinaria on the same bee legs, so your analysis would make it possible to find out whether the six pollinaria on the same bee belonged to the same species.  Of course, it would be best to test each pollinarium one at a time...right?



      On Thu, Aug 22, 2013 at 2:05 PM, Brosi, Berry J <bbrosi@...> wrote:

      Hi all,

      The Wilson et al. paper does not give a protocol for sequencing mixed pollen loads from bees.  It can only return one sequence.  Thus, it will give you a (single) sequence if there is a single pollen type in a pollen load, or a single sequence with some "indeterminate" base pair reads if there is a mixed pollen load with one species really dominating, or will return nothing if you have a very diverse pollen load (or if there is just a sequencing error!).  Such an analysis can work passably (marginally?) in a plant-species-poor area, especially when working with pollinators that are likely to visit a limited subset of plants, e.g. the Hawaiian Hylaeus that Wilson et al. studied.  But it is unlikely to work well in most other systems, e.g. which will have greater plant species richness and pollinators that are likely to visit multiple plant species.

      In addition, information on pollen quantities in a pollen load *cannot* be determined with molecular analyses currently.  DNA barcoding methods use sequences on organellar genomes (chloroplast, ribosome, etc.) because they are less subject to recombination than nuclear markers.  Copy number of organellar genomes is known to vary considerably—I don't know the specifics but I believe we are talking orders of magnitude—in pollen of different plant taxa, so even if one were to use quantitative PCR, etc., having greater relative quantities of PCR product of pollen from plant species A versus pollen of plant species B would not even necessarily tell you that species A has more pollen grains than B in your sample.

      Finally, unfortunately for most areas, there is very limited coverage of DNA barcoding markers for plants on GenBank, with the exception of a few areas that have been intensively studied for exactly that purpose.


      Berry J. Brosi, Ph.D.
      Assistant Professor
      Department of Environmental Studies
      Emory University

      On Aug 21, 2013, at 7:28 PM, T'ai Roulston <thr8z@...>



      I'm interested in knowing how you'll handle the quantitative question that Jack brought up. With Melissodes you should get a lot of unifloral pollen loads but are you going to examine pollen loads with a microscope to know the relative amounts collected when you hit mixed loads? Or are there good molecular protocols to handle this now?

      If you are still interested in Melissodes specimens, I could provide a few eastern ones.

      On Aug 21, 2013, at 5:46 PM, Karen W. Wright wrote:


      I will be trying to identify pollen from scopae of Melissodes. I am
      applying for a grant to help pay for ITS sequencing for the pollen. With
      the DNA you can measure genetic distance between host species which may be
      a good index of polylecty. If I don't get the grant, it is back to slides
      which is more time consuming, but cheaper, and less accurate (especially
      within Asteraceae). My biggest disadvantage is that I don't have a local
      library of ITS sequences since Melissodes' range is the New World. I will
      have to depend on genbank for sequence identification. I think we are
      getting to the point where there is enough available that this type of
      barcoding is possible. Any thoughts on this would be appreciated. Thanks,

      > Jack:
      > It's all relative. I still think that acetolysis is overkill (and often
      > counter-productive) if all you want to do is identify pollen carried by
      > the
      > animal visiting the flower vs. where the insect visited before it visited
      > the host flower. In my own case, I never count a pollen "species" taken
      > from an insect as "present" unless I find a minimum of 25 identical grains
      > or polyads (Linum, Acacia etc.) on the same slide. It's a trick I
      > developed in Australia. I put non-flower visiting flies and beetles in
      > the
      > same killing jar as my acacia-visiting insects and counted the number of
      > solitary grains the non-visitors carried due to contamination in the same
      > jar. The only "deal-breaker" should be the presence of a pollinarium
      > (orchid or asclepioid) provided you can find viscidium or corpusculum
      > still
      > attached to the insect.
      > Of course, comparing pollen DNA extracted from the bee's crop seems a bit
      > much but consider the future possibilities if the technology can be
      > tweaked. Eventually, it may tell you how many different trees (genotypes)
      > of the same species were visited by the same bee before it was captured.
      > Based on another Australian project I noticed that some pollen-swallowing
      > colletids invariably regurgitated their harvest if you killed them with
      > fumes of ether or ethyl acetate. Crop-derived grains are often a bit
      > "chewed up" but are still identifiable following staining with Calberla's
      > fluid.
      > "What are you doing, Professor?"
      > "I'm staining bee vomit."
      > "Oh... that's nice, dear." (It's always funnier with an Australian
      > accent).
      > Peter
      > On Wed, Aug 21, 2013 at 2:39 PM, Jack Neff <jlnatctmi@...> wrote:
      >> **
      >> The paper shows the method works but it seems like methodological
      >> overkill. I would think one could distinguish silversword (Asteraceae)
      >> and
      >> pukiawe (Ericaceae) pollen by much simpler and cheaper methods than DNA
      >> analysis. It is also not clear if this method gives quantitative
      >> results
      >> which is what one needs for diet analysis (not just presence or absence
      >> of
      >> various items).
      >> best
      >> Jack
      >> John L. Neff
      >> Central Texas Melittological Institute
      >> 7307 Running Rope
      >> Austin,TX 78731 USA
      >> 512-345-7219
      >> ------------------------------
      >> *From:* Colin Phifer <ccphifer@...>
      >> *To:* Bee United <beemonitoring@yahoogroups.com>
      >> *Sent:* Tuesday, August 20, 2013 7:57 PM
      >> *Subject:* Re: [beemonitoring] Identifying pollen collected from bees
      >> Another suggestion: you can now sequence the DNA for pollen ID. Here's
      >> an
      >> abstract from "Pollen foraging behaviour of solitary Hawaiian bees
      >> revealed through molecular pollen analysis" by Erin Wilson et al,
      >> published
      >> in Molecular Ecology vol 19 issue 21 pp 4428.
      >> "Obtaining quantitative information concerning pollinator behaviour has
      >> become a primary objective of pollination studies, but methodological
      >> limitations hinder progress towards this goal. Here, we use molecular
      >> genetic methods in an ecological context to demonstrate that endemic
      >> Hawaiian Hylaeus bees (Hymenoptera: Colletidae) selectively collect
      >> pollen from native plant species in Haleakala and Hawaii Volcanoes
      >> National
      >> Parks. We identified pollen DNA from the crops (internal storage organs)
      >> of
      >> 21 Hylaeusspecimens stored in ethanol for up to 3 years. Genetic
      >> analyses
      >> reveal high fidelity in pollen foraging despite the availability of
      >> pollen
      >> from multiple plant species present at each study site. At high
      >> elevations
      >> in Haleakala, pollen was available from more than 12 species of
      >> flowering
      >> plants, but Hawaiian silversword (Argyroxiphium sandwicense subsp.
      >> macrocephalum) comprised 86% of all pollen samples removed from bee
      >> crops. At lower elevations in both parks, we only detected pukiawe (
      >> Leptecophylla (Styphelia) tameiameiae) pollen inHylaeus crops despite
      >> the
      >> presence of other plant species in flower during our study. Furthermore,
      >> 100% of Hylaeus crops from which we successfully identified pollen
      >> contained native plant pollen. The molecular approaches developed in
      >> this
      >> study provide species-level information about floral visitation of
      >> Hawaiian
      >> Hylaeus that does not require specialized palynological expertise
      >> needed
      >> for high-throughput visual pollen identification. Building upon this
      >> approach, future studies can thus develop appropriate and customized
      >> criteria for assessing mixed pollen loads from a broader range of
      >> sources
      >> and from other global regions."
      >> I haven't tried this myself but it sounds very promising depending on
      >> your
      >> budget and timeline.
      >> • • • • • • • • • •
      >> Colin Phifer, PhD Student
      >> School of Forest Resources & Environmental Science
      >> Michigan Tech University
      >> 1400 Townsend Drive
      >> Houghton, MI 49931-1295
      >> Phone: (808)-315-2830
      >> Email: ccphifer@...
      >> On Aug 20, 2013, at 8:02 PM, Linda Newstrom <
      >> newstroml@...> wrote:
      >> Hi Laura
      >> We struggled with not wanting to take the time and expense to do
      >> acetolysis to identify pollen for our project (www.treesforbeesnz.org ).
      >> We are investigating the nitrogen content of pollen from bee-collected
      >> pollen pellets and have found that to get accurate identifications
      >> we really had to do acetolysis and what a difference it makes.
      >> But our project was focused on the pollen from pellets not a few grains
      >> on
      >> the honey bees.
      >> In any case, to identify unknown pollen a reference collection is best.
      >> We are creating a reference collection based on both the acetolysed
      >> pollen
      >> and a slide made of unacetolysed pollen from fresh or frozen pollen
      >> stained
      >> in fuchsin jelly
      >> from the flowers and matching them to the acetolysed pollen from the bee
      >> pellets for confirmation of identity of the pollen.
      >> The unacetolysed counterparts from the flowers show us how far we can go
      >> with not acetolysing in terms of making accurate identifications
      >> Depends on the number of species and size of location you are dealing
      >> with.
      >> When it is impractical to do acetolysis we have a reference collection
      >> with both views.
      >> We solved the practicality problem by partnering up with a palynologist,
      >> Ian Raine, at GNS Science, for identification of the pollen.
      >> The exchange of services and samples has been a win –win that serves
      >> both
      >> of our interests.
      >> See attached.
      >> Thanks
      >> Linda Newstrom-Lloyd
      >> *From:* beemonitoring@yahoogroups.com [mailto:beemonitoring@
      >> yahoogroups.com] *On Behalf Of *John Mola
      >> *Sent:* Tuesday, August 20, 2013 4:48 AM
      >> *To:* Doug Yanega
      >> *Cc:* Laura Russo; Bee United
      >> *Subject:* Re: [beemonitoring] Identifying pollen collected from bees
      >> I just wanted to back up a point Doug made: You only need a technique
      >> as detailed as the question you're asking. For me, I was going to avoid
      >> acetolysis but found it difficult to determine between Rubus and Malus
      >> pollens otherwise. If it had been simple to do that, or if I was asking
      >> a
      >> different question, it would have been fine to view the pollen
      >> untreated.
      >> Yes, acetolysis is rather time consuming and uses nasty chemicals, so if
      >> you can still make positive IDs without it, then go for it!
      >> On Mon, Aug 19, 2013 at 9:42 AM, Doug Yanega <dyanega@...> wrote:
      >> On 8/18/13 8:05 AM, Laura Russo wrote:
      >> Dear all,
      >> I am interested in identifying pollen samples collected from bees. I
      >> have
      >> looked at some resources on the internet, but I thought I would probe
      >> this
      >> listserve to see if anyone has had experience with that process, and
      >> therefore recommendations for resources (i.e. books, websites, people,
      >> etc.).
      >> Any advice on the collection methodology would also be gratefully
      >> received.
      >> I'll offer my own experience in two techniques, one of which is an
      >> example
      >> of the lowest-tech option (which I imagine many here will scoff at), and
      >> you'll have to determine for yourself whether either is a viable
      >> approach
      >> in your case.
      >> When I was interested in this type of analysis, I took a course in
      >> palynology, and found the techniques - all of which centered around
      >> acetolysis - to be time-consuming, somewhat costly (I was doing research
      >> out-of-pocket), and poorly suited for very tiny quantities of pollen.
      >> Kearns & Inouye was not yet in print back in 1983, so I was flailing
      >> about
      >> a bit and had to devise my own approach.
      >> (1) Working with pinned museum specimens of bees, the number of pollen
      >> grains available can be very small, and the technique I used was to take
      >> an
      >> SEM stub with double-sided sticky tape and scrape pollen onto it
      >> directly.
      >> While these samples could not be examined under a light microscope, and
      >> while they didn't have the outer surfaces completely clean, they could
      >> still be viewed under the SEM just fine, if I could find someone who
      >> would
      >> let me have a few minutes of SEM time (which was infrequent, so I didn't
      >> get to make much use of this).
      >> (2) for wild-caught bees (whose behavior I was observing) I used regular
      >> scotch tape, rubbed the bee gently against a small piece of scotch tape
      >> (technically, "magic transparent" tape which can be written on) to
      >> produce
      >> a pollen smear, released the bee, and stuck the piece of tape to a
      >> microscope slide. I wrote the ID of the bee and the date and time on the
      >> piece of tape in pencil. The pollen was viewed by flipping the slide
      >> over.
      >> These samples were untreated, and that has drawbacks, but it took less
      >> than
      >> a minute to gather a sample, from catching the bee to having the sample
      >> mounted and labeled.
      >> In both cases, some reference images or samples are needed, using pollen
      >> grains that have been *similarly untreated* (no acetolysis). Perhaps
      >> things
      >> have changed since I did this work (back in the 1980's), but I found
      >> there
      >> to be very few publications with SEM photos of raw pollen grains, and
      >> there
      >> was (and probably still is) absolutely nothing published showing raw
      >> pollen
      >> under a light microscope. In the former case, I ended up with many
      >> unidentifiable pollen types, but for my purposes it was still enough to
      >> know whether a given bee was foraging on multiple pollen sources or not
      >> (i.e., I had a very specific question, and knowing the plant ID would've
      >> been great, but not essential for those samples). In the latter case, I
      >> systematically went around in the habitat in question and took samples
      >> of
      >> pollen directly from the anthers of all the flowering plants in bloom
      >> within about 100 meters of the nest site to create a reference
      >> collection.
      >> This actually proved to be very effective, allowing me to identify the
      >> source of over 80% of the collected samples with an extremely minimal
      >> amount of time, energy, or expense - OR disruption of the behaviors I
      >> was
      >> observing (most females would, after a short interval, return to their
      >> nest, deposit the pollen load from which I had removed a smear, and then
      >> resume foraging).
      >> I haven't stayed abreast of recent advances, so can't speak as to how
      >> comparable my solution is to the methods used by others. I will note
      >> that -
      >> out of curiosity - I just pulled out my slides and looked at them under
      >> a
      >> scope, and aside from strong fading of the color, they look pretty much
      >> the
      >> same as they did nearly 30 years ago when I collected them. Scotch tape
      >> is
      >> about as low-tech as one can get, and it looks pretty good as a research
      >> tool. ;-)
      >> Peace,
      >> --
      >> Doug Yanega Dept. of Entomology Entomology Research Museum
      >> Univ. of California, Riverside, CA 92521-0314 skype: dyanega
      >> phone: (951) 827-4315 (disclaimer: opinions are mine, not UCR's)
      >> http://cache.ucr.edu/~heraty/yanega.html
      >> "There are some enterprises in which a careful disorderliness
      >> is the true method" - Herman Melville, Moby Dick, Chap. 82
      >> --
      >> John Mola
      >> JohnMMola@...
      >> ------------------------------
      >> Please consider the environment before printing this email
      >> Warning: This electronic message together with any attachments is
      >> confidential. If you receive it in error: (i) you must not read, use,
      >> disclose, copy or retain it; (ii) please contact the sender immediately
      >> by
      >> reply email and then delete the emails.
      >> The views expressed in this email may not be those of Landcare Research
      >> New Zealand Limited. http://www.landcareresearch.co.nz

      T'ai Roulston
      Curator, State Arboretum of Virginia
      Research Assoc. Prof., Dept of Envi. Sci.
      University of Virginia

      This e-mail message (including any attachments) is for the sole use of
      the intended recipient(s) and may contain confidential and privileged
      information. If the reader of this message is not the intended
      recipient, you are hereby notified that any dissemination, distribution
      or copying of this message (including any attachments) is strictly

      If you have received this message in error, please contact
      the sender by reply e-mail message and destroy all copies of the
      original message (including attachments).

    • Show all 16 messages in this topic