Automatic acoustic recorder is a low-cost, high-efficiency research equipment that can effectively monitor the activity level of species in a large space range and long-time span. To understand the rhythm of the Galliformes birds, vocalization of Brown Eared Pheasant (Crossoptilon mantchuricum), Koklass Pheasant (Pucrasia macrolopha) and Ring-necked Pheasant (Phasianus colchicus) were recorded using 40 automatic recorders in the Xiaolongmen National Forestry Park during 2019 breeding season (from April 27th to June 31st). The calls of Brown Eared Pheasant, Koklass Pheasant and Ring-necked Pheasant were extracted based on human-assisted machine learning. By optimizing the time-frequency parameters of the sound extraction, the correct rates of call extraction of the Koklass Pheasant and Ring-necked Pheasant were 73.32% and 89.32%, with the detection rates of 52.91% and 67.36%; however, the correct rate of call extraction of Brown Eared Pheasant was only 8.69% with the detection rate of 58.54%. Peak frequency, syllable duration, number of syllables and strophe duration were measured for each species. Acoustic characteristics were compared among species using K-S test and F-test. The calls of Brown Eared Pheasant were occurred rich in harmonics, while Koklass Pheasant and Ring-necked Pheasant were relatively few harmonics (Fig. 1, Table 1). Both breeding rhythms and daily rhythms on vocal activity were compared among species using T-test. In terms of breeding rhythms, all these three pheasant species had a peak of vocal activity during mating and early hatching periods (Fig. 2); in terms of daily rhythm, the Brown Eared Pheasant and the Ring-necked Pheasant had two peaks in the morning and evening, while Koklass Pheasant only had one peak of dawn calling (Fig. 3).