Or you could try identification of pollen grains by their shape and size.

Could you explain what exactly you want to identify? As to the question regarding crop preference: Do you want to know whether bees have some degree of preference towards certain plant species?

If you don't know what species to expect, you may collect pollen from bees and go for PCR with some plant-universal primers (we do this for animal prey with CO1 and/or 18S). Sequencing the amplicons (either upon cloning or with the aid of GS FLX or MiSeq Illumina) will tell you what was there.

Alternatively, you can use taxon-specific primers (e.g., of 18S or whatever seems suitable for your potential food sources) and analyse presence/absence of bands in agarose gels after nested PCR (universal primer first, taxon-specific PCR on the first PCR product)...

Dear Baharuddin, some techniques are available to extract DNA from pollen grains. For example:

-The determination of multiple microsatellite genotypes and DNA sequences from a single pollen grain. Molec Ecol. Y. MATSUK et al. 2007

- PCR and sequencing from a single pollen grain. Plant Molecular Biology. Petersen et al 1996

Or you could try identification of pollen grains by their shape and size.

Yes, David, of course, this would be much more straight-foward, easier and cheaper - if applicable (which I, being a animal ecologist, don't know)...

I agree with M. Zimmer and D. Inouye. The field of plant-pollinator networks has interesting approaches based on identification of pollen grains by their morphology.

I suggest you this interesting paper

Plant–pollinator networks: adding the pollinator’s perspective. Ecology Letters 2009. Jordi Bosch et al.

I think you don't have to use molecular methods to identify the food source of bees. For my as palynologist it is the most easiest thing to extract the pollen grains from the bees themselves or even from the honey and identify the plant species/gender/family from which they were collected.

I guess using molecular genetics methods is the hardest way to do that. Palinologists usually make a pollen reference collection with pollen taken from the plants in the area of the bee nests. They prepare slides after the cetolysis of the pollen grains (a process with allows a better visualization of its structure). Then the pollen from the nest is compared to the types of the collection, allowing you to determine more or less precisely from which flowers did it comes from.

Pollen identification by molecular methods is not really trivial. Bar coding like in animals is much more difficult for plants. One method could be AFLP investigations. A German written publication identificating Ophry pollen is: SCHIESTL, F.P. & VEREECKEN, N.J. (2008): Andrena combinata, ein neuer Bestäuber von Ophrys araneola. - J. Eur. Orch. 40 (1): 65 – 71. Another publication: WIDMER, A. & COZZOLINO S. & PELLEGRINO, G. & SOLIVA, M. & DAFNI, A. (2000): Molecular analysis of orchid pollinaria and pollinaria-remains found on insects. – Molecular Ecology 9 (11): 1911–1914. But you need a data bank for comparison. I think that traditional identification by shape and morphology is much more successfull.

Moreover, can you specify what is your sample? Pollen load, polled stored in the nest, pollen within the honey...

Thanks for all the answers and suggestions. Sorry, I am not familiar with both plant or bee, since my work is primarily on fish. Forgive my ignorance on knowledge of pollen morphology or botany. I am expecting something like barcoding or metagenomics (in animals for example identification of fish larvae) whereby we can take bulk sample of pollen and use Next Generation Sequencing. Can this method be done to identify pollen as well?

In this study, my collaborator took pollen load from the nest for identification of their compound based on the colouration. We hope to identify plant species they prefer (there are few bee species) by using the pollen.

Hi. Yes, using molecular methods will be a giod way to go but you do need good facilities. I suggest extracting DNA from pollen and amplify rRNA genes by PCR. Then clone and sequence or use something like tRFLP or DGGE to distinguish rRNA genospecies. Sequencing would be most informative and there are many rRNA sequences in online databases to help identify plant. ...but depends where you are and what ants are in the vicinity of bees.

BTW.You could use metagenomics to get whole genone sequence from pollen but this us unnecessary and added expense if all you need to do is ID.

IF you go for molecular tools (which, according to a couple of responses, may not be necessary), I suggest Next-generation-sequencing as the best choice: yes, it is expensive and not trivial(!), but with a single run you will get all the sequences present in your samples, and you can then compare them with sequences in the databases. By contrast, neither DGGE nor T-RFLP will provide pollen identity but just a profile of richness ("fingerprint") that you can compare among samples...

Traditionally, pollen analysis in apiculture is carried out according to LOUVEAUX et al. (http://www.airborne.co.nz/images/technical/methmels.pdf). You may also want to check the following:

http://www.tandfonline.com/doi/full/10.1080/00173130701760718#.UqG3DOKBaJo

http://www.ibra.org.uk/articles/20080612_40

http://www.ibra.org.uk/articles/20080611_109

An alternative method incorporates FT-IR spectroscopy (http://www.opticsinfobase.org/as/abstract.cfm?URI=as-57-1-23&origin=search).

In Allergome.org, you can see all the sequences of different pollens. So, you can use qPCR with these sequences.

In my laboratory we study concentration of different type of pollen using Flow Cytometry.

Dear Maria L. Espinazo Is possible to identify the specie/genus/family of a pollen grain using flow cytometry? I believe I am correct in saying that with flow cytometry you can assess the amount of pollen but no the identity.

If you have a good specific monoclonal or polyclonal antibody, you can identify correctly the genus. In the case of the species we must considerate the cross reactivity between them.

I send you a link of our last article about this topic:

"Determination of allergenic load and pollen count of Cupressus arizonica pollen by flow cytometry using Cup a1 polyclonal antibody."

"Cytometry B Clin Cytom. 2013 Aug 7. doi: 10.1002/cyto.b.21114. [Epub ahead of print].

http://www.ncbi.nlm.nih.gov/pubmed/23926112

If you can't download, say to me. It is so, I´ll send you e-copy the next Tuesday.

Given you are using pollen load, I think the best way is by using optical microscope. This will allow you to characterize the botanical origin and also having a relative quantification of the amount of pollen grain of different plants you will find. I think if you want to estimate insect preference, you also need to Quantify, not only to Characterize. Given this, all you need is an expertise technician, since different botanical species give different amount of pollen.. so the pollen count should be corrected according to that.

Thanks for all the answers and contributions. I am new in this field and I am actually a fisheries biologist. Cheers everyone!