Agrivoltaic systems, consisting of the combination of photovoltaic panels (PVPs) with crops on the same land, recently emerged as an opportunity to resolve the competition for land use between food and energy production. Such systems have proved efficient when using stationary PVPs at half their usual density. Dynamic agrivoltaic systems improved the concept by using orientable PVPs derived from solar trackers. They offer the possibility to intercept the variable part of solar radiation, as well as new means to increase land productivity. The matter was analysed in this work by comparing fixed and dynamic systems with two different orientation policies. Performances of the resulting agrivoltaic systems were studied for two varieties of lettuce over three different seasons. Solar tracking systems placed all plants in a new microclimate where light and shade bands alternated several times a day at any plant position, while stationary systems split the land surface into more stable shaded and sunlit areas. In spite of these differences, transient shading conditions increased plant leaf area in all agrivoltaic systems compared to full-sun conditions, resulting in a higher conversion of the transmitted radiation by the crop. This benefit was lower during seasons with high radiation and under controlled tracking with more light transmitted to the crop. As expected, regular tracking largely increased electric production compared to stationary PVPs but also slightly increased the transmitted radiation, hence crop biomass. A large increase in transmitted radiation was achieved by restricting solar tracking around midday, which resulted in higher biomass in the spring but was counterbalanced by a lower conversion efficiency of transmitted radiation in summer. As a result, high productivity per land area unit was reached using trackers instead of stationary photovoltaic panels in agrivoltaic systems, while maintaining biomass production of lettuce close or even similar to that obtained under full-sun conditions.