I'm sure you have this but - a blanket value for iops on a san is not helpful - you should break that down by lun/database and read writes. if you know the value has increased you should know where? the values you quote are very low anyway, my t-log drive does over 400 iops.
Anyway you should work out if this is reads or writes - technically you should have as few reads as possible as these would indicate physical reads whereas you really want reads to be from cache.
I have monitoring on each lun, it records reads, writes, write latency & read latency at 5 sec granularity, I store 6 months data and have summarised data ( 20 sec granularity ) for ever ( well since start ) this means I can track anything like this without too much issue. To be honest unless you have this type of monitoring you'll be blind and subject to every possible performance tuning suggestion in the world!
So points: - did you do any tests before the upgrade?
did you set a series of benchmarks to compare?
do you have historical performance data?
I can't say I've noticed any issues 2005 to 2008 / R2. As always the optimiser will be different so you should evaluate queries and indexes for chnages.
Have you checked ( here's the start of the list !! )
parallelism vs waits
any non out of the box settings
finally you say iops are up but how about performance? have you checked for suggested missing indexes and so on and so on.
I think you'll need to be a little more specific on your problems.
[font="Comic Sans MS"]The GrumpyOldDBA[/font]