Monthly Archives: January 2015

Another Day of Diving

Today Team Kelp went back to Bordjies in the Cape Point Nature Reserve and completed the sampling transects and quadrats that are part of Ross’ project. The wind started coming up early in the day so we had a decent bit of surge as we were only three metres deep, but the water was warm (17C) so the work went smoothly. No finger issues. Mostly sampling over crusting coralline with heaps of urchins. As I haven’t gotten gloves yet I had to remain aware while scraping the quadrats. Those little buggers are more than happy to remind you when you are in their personal bubble. Scraping the foliose algae off the rocks brought in a lot of klipvis. Almost manage to bag one but he eluded our grasp.

While the transects and quadrats themselves went well, finding the sites was a different matter. The center of each 8 metre radius site is marked out with an eye bolt that has been drilled into the rocks. These are then made more visible by having emergency tape tied around them. Problem is the cheeky little urchins like to use the tape as sunblock and so pull it down from the eye bolts and make a run for it with their ill gotten goods. Affirmative shopping isn’t restricted to humans. So in the future the plan will be to use brightly coloured rope that can’t be absconded with and that can be seen easier from the surface. Luckily the visibility was fantastic today. And, seeing how we were in the shallow kelp at Bordjies, we saw eagle rays, too đŸ™‚

Global Biodiversity Patterns

I started my PhD proposal presentation for UWC today. In which I will explain what I hope to accomplish with my dissertation to the Faculty of Science. Decided to use Prezi, as an almost complete split from Microsoft products is one of these intended goals.

The article for today was from Tittensor et al. (2010) in which they describe a massive planet-wide statistical analyses of species diversity and richness and how they correlated to six different hypothesis. 11,567 species across 13 taxa were included in this analysis. Divided broadly into coastal and oceanic species. The six variables were: SST, primary productivity, oxygen depletion, stable climate, supportive habitat and they accounted for different evolutionary histories across ocean basins. Unsurprising, though reassuring, warmer SST was found to have a significant correlation with increased species abundance and richness for all taxa (excluding pinnipeds) the world over. This applied to both coastal and oceanic groupings. After SST, beneficial habitat (e.g. longer coastlines or oceanic temperature fronts) were found to correlate well to areas of greater species diversity. Endothermic species (those that thrive in colder climates) like most pinnipeds, showed an increase in diversity/ abundance in areas of increased primary productivity. Ectothermic species did not share this correlation, though that would be expected as upwelling of cold water tends to be the primary driver of increased productivity.

The take away message for me from this paper was that yet again, a broad, well managed statistical analysis with an enormous abundance of data that tested multiple environmental variables found that SST is conclusively the primary correlate with diversity and abundance. Though they really should have used in situ temperatures for the coastal species. True story (Smit et al. 2013).

New Home

Hello World!

This is the first official post for me on this new site. Moving from Blogger to WordPress on this new domain. This also marks the beginning of me curating the domain. I’d spent the last couple of days struggling with FTP clients and reformatted my firewalls and routers in the process. No firm conclusion was reached other than that maybe the ISP itself is preventing the FTP clients from working… So I now have direct access to the cPanel software that AJ has already installed on the domain. This should allow me to access the domain in order to upload new data etc. The WordPress software is working (as evinced by the fact that this post exists) but am still learning cPanel. Haven’t actually tried to upload anything yet. Fingers crossed that it goes through…

Besides all of that fun tech stuff I put a couple of more hours into maths and a few hours into dissecting a long, detail and terminology rich paper on the effectiveness of different statistical tests at modelling species abundance and distribution (Meynard and Quinn 2007). The paper addressed the issue of terrestrial distributions but it seems like the same methodology would apply to marine ecosystems, too. Of the tests used, General Additive Models (GAM) came out on top with General Linear Models (GLM) a close second. It was also discovered that one is better off choosing the variables to be used in the models rather than letting the model work it out for itself. And not surprisingly the more abundant and well distributed a species is, the easier it is to model. Even under large sample sizes, if a species was rare (5% distribution), the variability produced in the model was always high enough to reduce the accuracy of the final product. I was a bit surprised to learn that additive models were generally more accurate than multiplicative models. This means that adding the effect the main variables would have on abundance is more accurate than multiplying them.

Anyway, that’s all for now.

More FTP fun, MEOW and R

So today marked round two of the FTP struggle. In order to access and customize the domain I need to be able to allow FTP through my firewall and router. This is problematic due to the default nature of these things doing the opposite of what I would like. It seems funny in a way that the main issue here is that FTP is too basic (developed in the 70’s before internet security procedures), it wants to make direct connections between client (me) and server (web domain) but the internet now-a-days is designed for almost all communications to go through a secure intermediary

After trying several different methods of allowing FTP through my firewall I decided that the issue must lay in my router. In an attempt to get into the router (which mysteriously was not letting me access it,even with a direct ethernet connection), I factory reset the little blighter. Unbeknownst to me, doing so requires a different login from our standard login to get the internet connection re-established. So I had to wait a few hours for afrihost to call back and sort out the issue. But that gave me plenty of time to read articles and finally get back on R.

Today’s article was by Spalding et al. (2007) and explained the new method of subdividing the coasts of the world into different ecoregions. It is a nested system with 12 Realms, 62 Provinces and 232 Ecoregions. A Realm is a large area, almost the scale of a continent that represents very a broad ecosystem that may span entire ocean basins. Provinces show a higher level of endemism with large groups of taxa only existing within a given province (the Benguela is a province). On the smallest scale, ecoregions represent areas along the continental shelf that may be more affected/ delineated by local biogeographical features (e.g. upwelling, freshwater input, rocky shores, etc.). The Benguela Province is divided into the Namib and Namaqua Ecoregions. The entire process they used to develop these Marine Ecoregions of the World (MEOW) was highly vetted and they received a lot of collaboration. MEOW is used by WWF and other conservancy organizations when they look at the necessity of certain programs depending on the region in question.

To round the evening out I finally dusted off my R software and got back in the programming saddle. It felt good to be doing something on a computer that actually responded to the inputs I was giving it. I had left off on a piece of code in October that was designed to plot all of our interpolated satellite data against all of our interpolated in situ data via a few easy to read windows. The data are all formatted to play nice, I just now need to decide how best to display the data. It is currently a big pile of madness. Need to delineate the climatolgies from each other better.