1 It seems to prefer areas that are what we might call "soft 



2 • substrate to the naked eye" but they're really very hard and there is a 



3 very firm attachment site, which is, I think, why it's restricted to 



4 canyons. 



5 This is Perimuricia grandis [phonetic] and an apathella, these are 



6 various corals that are found on hard substrates, these are the common 



7 ones. Here is Caraphenoides repestros [phonetic], a rattail that was 



8 found in significant abundances only in the canyon. 



9 This is a super-neat coral, Peregorgia arborea [phonetic], it's 



10 common in a lot of the canyons, all the way down to Baltimore Canyon. 



11 It's also common in one area up on the bank that's called "the trees." 



12 These gorgoneans grow to about 15 feet tall, and the reason it's 



13 called the trees is the fisherman don't like to tangle their nets in it, 



14 but a lot of juvenile fish like to hang around these corals, probably in 



15 terms of protection, a heterogenous environment. 



16 They were also common along the canyon axis, and here's just 



17 another picture. This is down at 1,200 meters, this is a cliff right at 



18 the base, the axis is down here. You can see all the corals, and then 



19 there are little brittle stars that are associated with the Perimuricia. 



20 Then another coral that was common along the axis on the cliffs, 



21 but also up on the flanks on boulders is Anthomastus agagazzi , also very 



22 much of a canyon indicator in that sense. 



23 Now, let's look at what this data looks like mathematically. 



24 Basically this is a slide showing what we did with the transects. We 



25 divided the transects up into 30-picture intervals, or intervals of 



26 different substrate. 



27 Then, what we did was we did community analysis and this is a 



28 percent similarity index. I don't expect you to interpret everything in 



29 here, let me just point out several things. 



30 You can tell the clustering structure is a function of depth, but 



31 it's also a function of location. The hot-pink areas, this is the 



32 number of areas. We started with a data matrix of 410 sample areas and 



33 141 species. 



109 



