PDA

View Full Version : Morphometry in a new light



gavin
27-10-2010, 01:16 AM
Well folks, Jon kindly sent me the raw data for colonies 44 and 46, two of his better ones. Curious to see how the data looked in 3 dimensions, I loaded three variables (CS, DI and HI) (OK then, CI, DS and HI) into a program written by Micha Bayer of SCRI's Bioinformatics group called CurlyWhirly. You can freely rotate the 3D plot manually by spinning the plot with the mouse. Here is one projection of the three variables for the two colonies. The two colonies fall into separate 3D space in some projections even though both are 'good' Amm. Do you like?!

418

The program can also take video clips of the rotating figure. Have done so, but it is taking hours to upload to YouTube so you'll have to wait for that.

G.

Jon
27-10-2010, 11:06 AM
I know a guy called Richard Bache who might be put into a spin by that programme!
Don't you just love it when beekeeping gets like Star Trek.

gavin
27-10-2010, 11:01 PM
Right! I boldly went where no man has gone before and finally converted 60% of the file (that's all the free download will let me convert) to a more suitable format and uploaded to your favourite video website.


http://www.youtube.com/watch?v=kEZ9XbWyFWM

It is more fun to download CurlyWhirly (http://bioinf.scri.ac.uk/curlywhirly/) and play with data yourself. The data files are plain text and there is an example which gives you the idea.

Jon
28-10-2010, 12:07 PM
How do you get the data from Drawwing into the application?

Hugh
28-10-2010, 01:27 PM
Maybe I’m too new to beekeeping to know what you’re on about and what it can do for you, but what do these graphs and video tell you and how can it help the poor bees??

Jon
28-10-2010, 02:21 PM
Hi Hugh.
It is a way of checking if you have native bees, imported bees, or some kind of hybrid between different races of bee.
Different races of bee can be distinguished by certain physical characteristics such as tongue length, colouration, or in the case of the graphs, characteristic patterns on the wings.
The native bee of the UK, Ireland and North West Europe is Apis Mellifera Mellifera.
Native bees are considered to be better adapted to our climate and foraging conditions, especially the further North or West you get, ie they can survive in the grim mizzle which we all know and love.
If you are interested in breeding AMM, the graph is a good indicator of the racial purity of your breeding stock.
The problem with bee breeding is that drones can fly for miles so the virgin queen can mate with a variety of drones of unknown provenance.
It should not of course be used an an exclusive means of selection of breeding stock as gentleness, honey production, frugal overwintering and the usual desirable features should be taken into account along with the graphs.

gavin
28-10-2010, 07:30 PM
Excellently put Jon.

To prepare the data for CurlyWhirly I added a column for 'Category' (colony), removed one dodgy data point, and pasted into a .txt file using Notepad. Then in CurlyWhirly you 'Open Data' under 'File'. Here are the contents of your wing morphometry file.

If anyone would like to take Jon's data (kindly extracted by Roger P) for a spin, copy the text below into a plain text file. As this site stripped out the necessary tabs between the data points, you will need to replace (<cntrl> H in Notepad) a single space with a tab (make a tab character in Notepad and paste the character into the 'Replace' command). For the morphometry freaks, I'd recommend trying this out as the spinning which I could capture in that video is constrained around the Y axis.

G.

Category Wing Ci DsA Hi
44 1 1.93 -5.348 0.9466
44 2 1.898 -1.921 0.8942
44 3 1.813 -3.788 0.9016
44 4 1.841 -5.733 0.8018
44 5 1.591 -4.783 0.8517
44 6 1.739 -4.373 0.923
44 7 1.458 -4.445 0.8378
44 8 1.626 -7.22 0.8715
44 9 1.596 -3.663 0.9639
44 10 1.693 -6.484 0.8528
44 11 1.527 -6.153 0.8676
44 12 1.492 -5.484 0.8346
44 13 1.684 -6.168 0.8281
44 14 1.837 -6.111 0.7802
44 15 1.624 -3.376 0.8317
44 16 1.927 -4.236 0.8966
44 17 1.787 -6.167 0.8114
44 18 1.548 -4.466 0.9412
44 19 1.571 -4.276 0.9052
44 20 1.61 -2.979 0.8566
44 21 1.497 -3.82 0.8944
44 22 1.53 -3.025 0.9494
44 23 1.908 -4.455 0.8594
44 24 1.752 -5.09 0.9206
44 25 1.463 -3.457 0.876
44 26 1.759 -7.427 0.8004
44 27 1.673 -5.623 0.8817
44 28 1.717 -4.821 0.806
44 30 1.791 -2.559 0.8741
44 31 1.912 -3.924 1.0016
44 32 1.812 -1.665 0.9489
44 33 1.783 -5.225 0.9485
44 34 1.524 -5.754 0.8139
44 35 1.826 -3.754 0.9923
44 36 1.91 -3.794 0.8504
44 37 1.819 -3.06 0.8456
46 1 1.716 -7.012 0.8604
46 2 1.851 -3.381 0.7668
46 3 1.673 -4.963 0.9029
46 4 2.025 -5.744 0.8084
46 5 2.004 -3.002 0.861
46 6 2.025 -3.082 0.9152
46 7 1.858 -2.818 0.8837
46 8 1.728 -3.726 0.8613
46 9 1.821 -1.658 0.8944
46 10 1.614 -5.161 0.7797
46 11 1.603 -2.075 0.8652
46 12 1.635 -2.144 0.8479
46 13 1.753 -3.834 0.8138
46 14 1.663 -2.277 0.9176
46 15 1.78 -5.713 0.7945
46 16 1.515 -5.32 0.7831
46 17 1.951 -2.256 0.8011
46 18 1.817 -2.917 0.7806
46 19 1.702 -3.734 0.8202
46 20 1.768 -4.761 0.8724
46 21 1.548 -2.709 0.8173
46 22 1.816 -2.678 0.9175
46 23 2.062 -3.638 0.858
46 24 1.524 0.9758 0.9386
46 25 1.693 -1.671 0.8534
46 26 1.593 -2.167 0.8806
46 27 1.297 -5.074 0.7462

Jon
28-10-2010, 08:05 PM
I was hoping you would have come up with some way to take the data straight out of the Drawwing excel file.
I'll give it a go with some of the other results I have.

gavin
28-10-2010, 10:15 PM
You just copy and paste from Excel to Notepad, and you're done. To plot more than one colony just add a column at the start, insert the name you are using for the colony, and then add extra colonies as extra rows.

Jon
28-10-2010, 10:19 PM
Well I gave it a go. Not sure what I learned but it was fun standing on the bridge of the Enterprise for a minute or two.
If you put in data from two similar colonies it just rotates in a similar fashion - not surprisingly I suppose.
We need a volunteer with a set of Buckfast or Carnica wings to expose to the dilithium crystals.
I remember Roger had scattergrams from a few diverse colonies and swarms which were published in the bibba magazine a few months back.
Those might be interesting to plot against native type bees.

Jimbo
28-10-2010, 11:36 PM
Calling star ship Enterpise. I have a set of Carnica results taken in 2007 AD using the Beemorph software. How do you want me to teletransport the results

gavin
29-10-2010, 12:05 AM
Captain's Log 13.4.21. On Planet SBAi we encountered a novel life form with unusual banding on the abdomen, and overall dark appearance, a slim outline and calm demeanour but a propensity to eat up all their provisions as soon as they brought them in.

Beam it over any way you like Jimbo. Paste it into here, email it as whatever file you have, paste it into a PM ....

There was already a possibility of a digital slide show on 27th of next month at Scotlandwell. It is starting to look like spinning carnies and Amm wings will be on the cards too.

Muchos gracias

Gavin

Jimbo
29-10-2010, 09:11 AM
Will do Captain. Don't have the original wing scans but the CI and DS results for each wing from the Beemorph program which I then used in the Drawwing program. It gave the same result. I will e-mail it to you tonight as an attachement.
The queen was originally imported from Denmark and I was told it was a Carniolan. The history of this queen is interesting. Year 1 great gentle queen lots of honey. Year 2 same queen gentle and lots of honey. Year 3 new queen produced and mated with local bees gave aggressive bees and less honey. Year 4 requeened with a local queen. The drones from this original queen I can still pick up in the morph plots especially in the apairies near to the original site. It has taken about 3 -4 years for the Carniolan plots to be diluted out in the local bee population and this was only from one queen

Jon
29-10-2010, 10:37 AM
Carnica is the problem because they are dark and the drones can sneak under the radar. The Buckfast or Ligustica crosses show up easily because of the yellow banding.
I don't even test colonies with a lot of yellow bees as I would not breed from them. They can be fine for making drones though if you are sure about the origin of the queen.
That's where morphometry comes into its own as Drawwing should pick up Carnica crosses easily due to the higher than expected CI values.

Both the plots below were very black bees, the first a swarm and the second from my father's apiary. You would be hard pushed to tell them from my bees on appearance.

422 421

The bees from plot B were as runny on the comb as I have ever seen and those from the swarm were aggressive. I have noticed with my father's bees that they look superficially like mine but his queens are different - not as black and with more banding.

gavin
29-10-2010, 12:45 PM
That's an interesting tale for the 27th Jimbo.

Do you also have the Hantel Index? Unfortunately you need a third variable for a 3D plot, but even in 2D it should be interesting.

G.

Jimbo
29-10-2010, 03:52 PM
Hi Gavin,

The original wings were done using beemorph. I don't think beemorph gives the Hantel. If you take the beemorph plots and substitute them into the Drawwing software it only gives you the CI /DS ratio and not the other ratio's ie Ci / Hantel etc that a normal scanned Drawing result would give, but there may be a way to work it out.
When I was on the BIBBA Morphometric trainer for trainers course I got the impression that the Hantel values were not that important to the results obtained by Drawwing.
One way to get the Hantel index from this data may be to use another program ie I have seen C-Wing but have no experience on how it works.
Will send all data I have got tonight as an attachment.
The Danish queen I was informed was Carniolan by Ben but when I was in Denmark I thought it was a form of Buckfast/Carniolan that they were breeding. I can send an e-mail to Denmark to confirm the type. There is certainly no trace of black bee whatever it is.

Neils
29-10-2010, 03:56 PM
Should be able to get you some random samples of "non native" bees if you want some. Can do you pretty dark, kind of dark witgh some banding and very yellow bees (don't you just love technichal terms).

Jimbo
30-10-2010, 10:29 AM
Hi Nellie,

If you want to get more technical send me a samples of your bees I will do the Morph plots for you and let you how native they are. E-mail me for instructions on how to take the samples and for sending them to me.

Jimbo

beebreeder
30-10-2010, 10:15 PM
Hi All
How good a scanner do you need these days for plotting, and what is the best s/ware to use. Our association looked into it years ago when we had some mc killlocody queens, but due to bad matings the temper just got worse, the association apiary is back to local mongrels now and reasonable temper, I have gone my own way and raise my own queens.
kev

As requested I've moved some posts to a new thread discussing Morphometry Procedures and Standards (http://www.sbai.org.uk/sbai_forum/showthread.php?349-Morphometry-procedures-and-standards). Perhaps a software discussion belongs there too.

G.

Peter
04-11-2010, 01:20 AM
I did it this way:
In MorphPlot insert a new sheet and head column A 'Category'.
In the MorphPlot sheet select the data and headers, i.e. the cells with Wing, Ci, DsA and Hi, and all the data in the columns below the headers.
Paste this into the new sheet in cell B1.
If you wish to inset a 'Category', enter this in A2 and copy it down (select it and the cells below and press Ctrl+d.
You now have all the data required in columns A to E, but it will need saving as a text file and the spaces replacing with tabs. I gave up using notepad several years ago in favour of EditPad Lite ( free from http://www.editpadlite.com/).
Select and copy all the data.
Open EdipadLite and paste the data into it.
Click on Search/Show search panel.
In the search box type a space and in the replace box type a tab. You will not see them but they will be there.
Click on Replace all. Your data will now be tab separated.
Click 'File/Save as' and save the file with the name that you wish to give it.
Open it with CurlyWhirly.