Contrast sensitivity in pigeons: a comparison of behavioral and pattern ERG methods

Contrast sensitivity (CS) is often used to assess spatial and temporal vision in animals. Conventional behavioral psychophysical techniques are both time and labor intensive, whereas measurement of CS functions by means of the pattern electroretinogram (PERG) is considerably more rapid and efficient. Are the two methods comparable, however? To answer this question, contrast-sensitivity functions were obtained using both the PERG and behavioral psychophysics in the same subjects, which were White Carneaux pigeons. The stimuli, in both methods, were phase-reversing, contrast-modulated sweeps of sinusoidal gratings. The PERG-CS functions were recorded via corneal electrodes and the behavioral data were collected using a modified staircase method that used moderate food deprivation and food reward. The results indicated that the PERG-CS functions had comparable bandwidth and peak spatial frequency to the behavioral CS functions. The PERG-CS functions, however, were lower on average than the behavioral curves by about 54%. The visual acuity of the two methods, as estimated from the high-frequency cutoff of the CS functions, differed by 37%. Both of these values are roughly consistent with the square root of 2 advantage of binocular viewing (behavioral method) over monocular viewing (PERG method). In addition, the peak spatial frequency showed a decrease of 0.125 c/deg with the PERG method and bandwidth was reduced by approximately 0.5 octave. These findings suggest that the PERG is an acceptable alternative to behavioral measurement of CS functions, especially in animal psychophysics, if one takes into account the underestimation of CS by the PERG method and the small changes in peak spatial frequency and bandwidth.