Another Kiwi! baldtankman aka Mark

Discussion in 'Introduce Yourself' started by baldtankman, Nov 13, 2008.

  1. baldtankman

    baldtankman Wonderpus Supporter

    Joined:
    May 5, 2008
    Messages:
    184
    Likes Received:
    0
    Hi,

    SOS pointed out that I have never introduced myself so here goes.

    Hi all I am baldtankman aka Mark Fenwick. Until recently I worked at Te Papa as a tech, to help fund my studies at Uni.

    I worked with Bruce Marshall on a few things, (he basically dragged me through my thesis, the genetics of the NZ freshwater mussels) one of those being giant squids. Bruce introduced me to SOS and I have had a great time with Steve and Kat on a few projects over the years.

    Anyway I now work as a benthic ecology tech at NIWA the NZ national institute of water and atmosphere where I hope to see a lot more Ceph's and maybe do some more work on squids!

    Cheers
    Mark
    :yinyang::talker:
     
  2. monty

    monty Colossal Squid Staff Member Supporter

    Joined:
    Mar 8, 2004
    Messages:
    4,887
    Likes Received:
    11
    You mentioned in another thread that you did "fourier curve analysis" on mussels... I know what fourier analysis is to some extent, but not how to use it for 3d or 2d curves like mussel shells, so I'm curious about the details? Do you transform the shell's shape into polar coordinates and then do fourier analaysis on the radius, or something like that? Or do a 2-d FFT on a picture of the shell? or something else entirely?
     
  3. Steve O'Shea

    Steve O'Shea Colossal Squid Supporter

    Joined:
    Nov 19, 2002
    Messages:
    4,668
    Likes Received:
    17
    :smile: So, you're branching into squid Mark. Good choice. If you're ever out on the boats I'm mad keen on getting any Teuthowenia (cranchiid) that comes in, regardless of its condition. Jens Horstkotte is (re)joining us in January '09 to do his PhD on the group, starting in NZ and then moving further afied (southern oceans stuff), systematic review. (His Masters was on histioteuthids.)

    Times are good down here in lil' ol' NZ, as far as systematics of cephalopods is concerned.
     
  4. baldtankman

    baldtankman Wonderpus Supporter

    Joined:
    May 5, 2008
    Messages:
    184
    Likes Received:
    0
    Hi Monty,

    "Do you transform the shell's shape into polar coordinates?" Yes:smile:

    We are branching into a less derived taxa indeed. The basics are that we are using 2d digital images of a right valve that we have photographed on a light table (but any good image works). We trace the outline of the shell using something like imagej then derive xy co-ordinates of the outline. Then we have a clever little program (HTREE) that normalizes the outline for size and introduces a bit of "noise" to allow for differences in the starting point (important in bi-valves as the umbo is the only consistent character). So then the program does its bit on the data and produces a data set that we do a PCA on in SYSTAT or Sigma.

    James Crampton at IGNS Wellington knows much more about the stats than I do and he is very approachable if you want more details. He is interested in comparing my data set with his as mine is calibrated with DNA and he works on fossil bi-valves.

    I have been wondering if we could use this sort of approach on squid beaks, maybe using landmarks rather than Fourier.

    Cheers
    M
     
  5. baldtankman

    baldtankman Wonderpus Supporter

    Joined:
    May 5, 2008
    Messages:
    184
    Likes Received:
    0
    Yes we certainly need to have a chat. I might be able to be bought!:yinyang::smile:
     
  6. monty

    monty Colossal Squid Staff Member Supporter

    Joined:
    Mar 8, 2004
    Messages:
    4,887
    Likes Received:
    11
    cool stuff. PCA is a great algorithm.

    If you do the squid beaks, you might want to talk to Hallucigenia, who did some work ( http://pda.physorg.com/lofi-news-condor-species-syverson_112891586.html ) using PCA analysis on some measurements of condor bones, which I think involved landmarks and distances between.

    Something like this seems potentially interesting for ammonite sutures and shapes, too.
     
  7. baldtankman

    baldtankman Wonderpus Supporter

    Joined:
    May 5, 2008
    Messages:
    184
    Likes Received:
    0
    For sure once you have a system set up its a pretty easy way of splitting species..
     

Share This Page