Thanks MattW, I hadn't thought of the Canadian universities before. From the website, it doesn't look like they require the Math GRE, but do you think I ought to send results to them anyway (I'll be taking it in April)? You don't think that my lack of coursework on my transcript is going to be a problem at a lot of these places?
I'm definitely much more of a visual person so I think anything in analysis or topology would suit me. I think a field like differential geometry or complex analysis would be a good fit for me, though I'd still like to cast a wide net.
Anyhow thanks for all the help. Keep the suggestion coming! You guys are definitely helping me to firm up a picture of the types of places I can expect to be a competitive applicant!
I wouldn't bother with sending the Math GRE to Canadian universities. It's unlikely to affect your chances of being accepted unless the score is exceptionally good or bad.
It sounds like Toronto or UBC would be a good fit for you, although both are likely out of your reach. I think that I will keep my recommendations of Waterloo and Alberta, and throw in the recommendation of Western Ontario. It looks like Western Ontario is most in line with your interests. Waterloo and Alberta are both stronger and larger programs but lack anybody working in complex analysis. However, Waterloo does have a strong analytic number theory group which involves a ton of complex analysis. Look up, for instance, modular forms.
Both Alberta and Waterloo have super analysis research groups, with members working in a large number of different flavours. In particular, Waterloo has the strongest functional analysis group in Canada (which is my research area). I can't comment on the differential geometry/topology group at Alberta but the geometry/topology group at Waterloo consists of young and very active members.
In the interest of full disclosure, I'm a PhD student at Waterloo. If you want any more information about Waterloo, send me a PM. For instance, I would be happy to point out potential supervisors.