To better assess ecosystem C budgets of croplands and understand their potential response to climate and management changes, detailed information on the mechanisms and environmental controls driving the individual C flux components are needed. This accounts in particular for the ecosystem respiration (Reco) and its components, the autotrophic (Ra) and heterotrophic respiration (Rh) which vary tremendously in time and space. This study presents a method to separate Reco into Ra [as the sum of Ra (shoot) and Ra (root)] and Rh in order to detect temporal and small‐scale spatial dynamics within their relative contribution to overall Reco. Thus, predominant environmental drivers and underlying mechanisms can be revealed. Reco was derived during nighttime by automatic chamber CO2 flux measurements on plant covered plots. Rh was derived from CO2 efflux measurements, which were performed in parallel to Reco measurements on a fallow plot using CO2 sampling tubes in 10 cm soil depth. Ra (root) was calculated as the difference between sampling tube CO2 efflux measurements on a plant covered plot and Rh. Ra (shoot) was calculated as Reco – Ra (root) – Rh. Measurements were carried out for winter wheat (Triticum aestivum L.) during the crop season 2015 at an experimental plot located in the hummocky ground moraine landscape of NE Germany. Reco varied seasonally from < 1 to 9.5 g C m−2 d−1, and was higher in adult (a) and reproductive (r) than juvenile (j) stands (g C m−2 d−1: j = 1.2, a = 4.6, r = 5.3). Observed Ra and Rh were in general smaller compared to the independently measured Reco, contributing in average 58% and 42% to Reco. However, both varied strongly regarding their environmental drivers and particular contribution throughout the study period, following the seasonal development of soil temperature and moisture (Rh) as well as crop development (Ra). Thus, our results consistently revealed temporal dynamics regarding the relative contribution of Ra (root) and Ra (shoot) to Ra, as well as of Ra and Rh to Reco. Based on the observed results, implications for partitioning of Reco in croplands are given.