Here's a PAR blurb from Andrew Trevor-Jones
PAR is a measure of the amount of light reaching a surface rather than the output of a lamp. i.e. the number of photons hitting an area over time. As such the PAR of a lamp is undefined.
PAR also treats all photons equally. Photons of a higher shorter wavelength (higher frequency) have higher energy levels than longer
wavelength photons. The differences in energy levels is not taken into consideration with PAR - although a meter may have to deal with it and normalise the readings.
Metal halide lamps are point light sources and so the further you move away from the lamp, the greater area over which the light is spread and so the lower the irradiance. If measuring PAR of a metal halide lamp, you will get different readings at different distances from the exact same lamp. You will only be able to compare the PAR from the different lamps if you are measuring PAR under the exact same conditions - i.e. the exact same distance from the lamp with the lamp in the same fitting and reflector. Even with these things being the same, there may be subtle differences in the readings due to aberrations in the glass of the lamps.
Another thing to consider is the PAR meter being used. Does it measure all wavelengths equally? The PAR meter I have (from Apogee Instruments) under reads blue light and over reads red light. If you are using a LiCor sensor, this may be less of a problem, but it is worthwhile finding out the spectral sensitivity of the sensor.
Yet another factor to consider is that different brands of lamps will have different output spectra and light output even if they are the same wattage and claimed colour temperature. You may even see differences between lamps of the same brand and same claimed colour temperature. Also,the claimed colour temperatures of lamps is only very approximate and gas discharge lamps don't really simulate theoretical black bodies.
Assuming that you are measuring each lamp under the same conditions and the sensor is not under reading blue light, you should still expect to see less PAR with lamps of the same wattage but with a higher colour temperature because there will be proportionally more photons with higher energy levels. As energy can not be created nor destroyed, the same amount of input energy (the wattage of the lamp) cannot produce more energy and so there will be fewer total photons produced by a lamp that
has more blue light.
If the new lamp looked more blue then it would have had a higher approximate colour temperature. Whether it was 20000K and the originals were 10000K is not going to be easy to measure. If you could measure the spectrum of each lamp, you could compare that to the spectra of theoretical black bodies at various colour temperatures.
I hope this helps.
Andrew Trevor-Jones
Now I recently read that cyano love the yellow-red spectrum, a portion of the spectrum not high in the LED output, so LED's theoritcally help eliminate cyano.
I read that soemwhere and then went to the source study which does seem to state just that.