This study was set to investigate the effects of photovoltaic(PV) arrangements on photosynthetically active radiation(PAR). Field experiments on tea plants, an important economic crop, were conducted during their peak growing season(June to August) within PV arrays of an agrivoltaic base. Daily and hourly average solar radiation beneath PV systems with different configurations was simulated using Ecotect software. Light response curves of tea plants were measured with a Li-6800 portable photosynthesis system and fitted using four photosynthetic models. The photosynthetic rates of the tea canopy under PV systems were then predicted by combining these fitted models with measured under-panel radiation. The results showed that the non-rectangular hyperbola model provided the best fit, exhibiting the highest coefficient of determination(R2=0.997 9), lowest mean squared error, and lowest Akaike information criterion. Model estimates of key photosynthetic parameters including maximum net photosynthetic rate, light saturation point, and light compensation point matched observations. Under agrivoltaics, daily average photosynthetic rates declined by 3.40%-8.07% and hourly average rates between 09:00 to 11:00 by 2.97%-6.90%. The strongest inhibition on tea plants photosynthesis occurred with 20° tilt, 3.9 m height, and 8 m spacing; the weakest with 28°, 2.5 m, and 10 m.