Thursday, 23 August 2012
EMS Project: Mapping bird distribution.
I feel like I have completely abandoned this blog - I haven't, I've just had my mind in other places (i.e. stuck in ArcGIS, my old pal) and having a little bit of me time with my other blog. I've had a lot of things on my mind and, without wanting to clog up this blog or my main Twitter account (RachLinaker) with personal-rubbish, I put it onto my other. Now, everything is all right in the world - woohoo.
So, of late I have been inputting the reams and reams of data collected by those lovely people of the EMS management team (massive thanks to you!) and looking at the survey sheets that go alongside them to gauge where abouts on the shore the birds were recorded from. It has been one heck of a task and one that took much longer than anticipated, as such I am now crying out for 'the fear' to hit me. Come on... Please! Anyway, I'm finally done - hurrah.
Things were looking a little messy (trust me, when you're playing around with hundreds of points, the little colourful balls on the page do resemble a child's birthday party... I've said that one before) so I wanted to play around with the shapefiles and have them represent high and low tide, great, except some sides were only sampled and high or low. I scrapped that idea and hopefully, once I've done the statistics, this will back it up (I've saved the new layers just in case there is a significant difference between either numbers or activity patterns - probably a little difficult to do with a relatively small sample size). Instead, I opted to divide the counts up as to whether they were feeding or roosting - this should give an idea of what areas of each site are used more readily by each species and give an idea of these little hotspots so to speak. Looking back, this is exactly why the mapping was done in the first place so that's good. I've come across a few bumps along the way unfortunately, one being that for some sites, activity just was not recorded. Others, for example with Bran Sands in particular, multiple birds were noted on the survey sheets but numbers hadn't been noted beside them so it may be a little skewed. Also, I've had to scrap the "end count" data with regards to the mapping - there were just too many that either weren't counted or activity was accounted for or anything like that. Not like anybody cares about this but it's a nice note for me at least!
I still need to sort out the legend somewhat. I've probably made things very difficult for myself, but essentially the legend is only showing the size classes used for the map - fine - except I've made two separate layer files so that I can show both Feeding and Roosting birds, and then only select one or the other - now GIS is creating duplicate birds on the legend... I think I may have to manually manipulate it since my brain is officially dead tonight.
What I need you to do is imagine that the legend looks awfully pretty. I think I'm going to have another play around and set it out on portrait paper, not landscape, and then at least the legend can go down vertically on the right. At least you can see that certain species do prefer certain areas of the habitat...
I'm going to have a google around and see about placing two different maps on the same page. There must be a way of doing it and I'm just being a little bit blind.
More to come!