Carbon cycling in terrestrial ecosystem is of great importance in global change issue. A reliable estimation of plant respiration
is essential to objectively evaluate the role of vegetation in the terrestrial ecosystem cycling. As an easily measured surrogate for protein content, nitrogen concentration plays a vital role on plant respiration. The dependence of plant respiration on nitrogen concentration for forest ecosystem has been well pronounced over the last decades. However, few studies have been dedicated to agro-ecosystem. To quantify the dependence of crop respiration on tissue N, crop growth and temperature, we conducted pot and field experiments during 2002 and 2003 wheat- and rice-growing seasons with different treatments, including fertilizer N application, planting date and planting density. Crop biomass, tissue N content,
dark respiration and the corresponding environmental temperature were measured. Static opaque chamber was used for sampling gas. The respiration as CO 2 emission was detected by a gas chromatograph (Agilent 4890D) with flame ionization detector (FID). A gradual-clip-up method was employed to determine the dark
respiration coefficient (R d), defined as the CO 2 emissions per unit crop biomass per unit time at a reference temperature. Results indicated that the R d of wheat and rice crops is not a constant but is positively dependent on shoot N content. Relationship between these two parameters can be well quantified by R d=4.74N-1.45 (R~2=0.85, n=122, p<0.001). The lowest and the highest values are 3.58 and 31.49 μmol kg~(-1) s~(-1) for the wheat crop over the 2002-03 growing season, and 1.04 and 29.56 μmol kg~(-1) s~(-1) for the rice crop during the 2002 growing season, respectively. Crop dark respiration (RD) associated with tissue nitrogen (N), crop biomass (W) and air temperature (T) can be quantified by RD=(4.74N-1.45)×Q~((T-25)/10) (10)×W. Q (10) is a temperature coefficient for the respiration. Computed RD applying this relationship matched field measurements well. We computed daily crop respiration over a rice-growing season and a wheat-growing season in Nanjing by using the methods in the crop model, the TEM as well as derived from this study, respectively. Calculations indicate that the estimated respiration with crop model is in general comparable to this study, while the respiration estimated by the TEM method is 59% (for rice) and 31% (for wheat) higher than that from this study. It is concluded that the tissue N content affects greatly crop dark respiration. The respiration coefficient is linearly correlated to tissue N content. Crop dark respiration can be well estimated from tissue nitrogen, crop biomass and air temperature. A further conclusion is that the simulated dark respiration of rice and wheat crops is comparable to the maintenance respiration estimated by the crop model, while the TEM method overestimated crop maintenance respiration and may be not suitable for agricultural vegetation.