Generally speaking compounds that absorb in the UV region have relatively broad peaks. It is normal practice for quantitative HPLC analysis using a UV detector to detect the peak of interest at its maxima. If this is not possible you have to take account of the fact that the wavelength of a UV detector varies with temperature. So if you are doing quantitative analysis of a compound at a wavelength other than at its maxima (on a slope of the UV curve) then if the laboratory temperature changes so will the peak area. This has large quantitative effects if you are doing analysis of several peaks in a chromatogram with different UV maxima.
Harshad are you asking about wavelenght or detector speed?
Everybody here focused on the first one and they gave you good answer, now I'm focusing on the second one.
It could affect your peak response: if you choose a too slow acquisition frequency (e.g. 1Hz for a peak 5-6sec wide) you could miss you most intense acquisition point and so you'd obtain a false low peak area.
On the opposite side if yuo set your PDA to 10Hz for peaks around 15-20sec you'd obtain a noisy signal.
If you set your PDA speed to obtain 15-20 points across an "average" peak of your chromatograms you'd work in the optimal conditions.