[ChrisK]

What I'm not getting is why the PAR curve doesn't drop off in the green part of the spectrum which is inefficiently used for photosynthesis.

edit: Here is the PAR curve I am familiar with and it's significantly different from above:

http://www.tomatoville.com/showthread.php?t=26446

I assume the first graph depicts photosynthesis rate (No Y axis labels) at each wavelength.

Originally Posted by sio2rocks

I'll Answer for hotwired since he isn't here right now. The PAR curve is the region of visible light between 400 and 700 nm emitted by our Sun that plants can use for photosynthetic processes. PAR stands Photosynthetically Active Radiation. The curve shows radiation energy in some units on the Y-axis that aren't important for our purposes. On the x-axis in the wavelength of light that those energies occur along the curve. So if we want a light source that is the perfect clone of our Sun we would want its radiant energy to fall exactly on the PAR curve shown in the graph below.

However no (cheap) artificial light source by itself can do this so we opt for one that replicates part of the spectrum well and has some issues with other parts. For vegetative growth the lower regions of the PAR spectrum are of most importance (say between 400 and 550 nm), mostly blue and green light spectra. For flowering/blooming plants we would want more yellow, orange, red light (550-700 nm) so we would try to replicate the higher range of spectra for them. A typical "daylight bulb" has a color temperature of 6500k (term for total radiant energy from light related to temperature in Kelvin, based on the radiant energy from a black body). This bulb has the radiation breakdown in the visible spectrum (400-700 nm) shown below next to the PAR curve.