Thanks for the answers.
I'm going to try to explain my problem, it's very hard to me to write in English (I'm Spanish) .
I have a big spritesheet with plants, but in each level I only use one or two, so i think it will be better to have each plant in a different small file. If i test it with my current levels there are no performance changes apparently, because it runs always at 60fps. But for example, let's think the game (without the 60 fps iphone limitation) would run at 120 fps and with my new changes at 60 fps, i wouldn't notice the different because i always see 60 fps.
I think NSTimerInterval or CCProfilingTimer won't solve my problem because i don't want to test how fast is a peace of code in terms of CPU but the improvement of fps because of the change in my textures (maybe now opengl has to do more swaps with my textures o maybe now there are smaller textures and runs better).
Maybe if i use [[CCDirector sharedDirector] setAnimationInterval:1.0/120]; and check the delta is a solution but i don't know how to do it in an easy way (without using NSLog that affects the performance).