Electromagnetic precursors have been investigated for a recent relatively large earthquake (EQ) with magnitude M=6.7 occurred in the Hokkaido Island, Japan on 5 September (UT), 2018 on the basis of coordinated data on the lower ionospheric perturbation with subionospheric VLF/LF propagation anomalies monitored in Japan and in Russia, on the upper F region anomaly with GIM (global ionosphere map) TEC (Total Electron Content) and on the stratospheric behavior with the use of ERA5 temperature profile data. It is found that there were observed the clearest VLF/LF propagation anomalies indicative of seismogenic lower ionospheric perturbations above the EQ epicenter on 4 and 5 September before the EQ and 6 September even after it because this period was geomagnetically quiet. As for the F region, an anomaly in TEC over the EQ epicenter appeared only on 5 September, but it is quite uncertain whether this anomaly is related to either a pre-EQ effect or a geomagnetic storm effect. Then, we have tried to find if there were some seismo-related AGW (atmospheric gravity wave) activities in the stratosphere, but though we have found a lot of anomalies, they are highly likely to be associated with the activities by convective weather systems. Even though we have recently identified the first convincing direct evidence on the Atmospheric oscillation hypothesis in the lithosphere-atmosphere-ionosphere coupling (LAIC) mechanism for the 2016 Kumamoto EQ [Yang et al., 2019], this paper aimed initially to obtain further evidence on this channel. However, the above experimental findings may provide a negative result in this direction; that is, this Atmospheric oscillation channel is unlikely to be in operation for the present 2018 Hokkaido EQ, suggesting other channels in the LAIC effect. So we have made extensive discussions on this LAIC mechanism for this Hokkaido EQ with special reference to the 2016 Kumamoto EQ and the 2011 Tohoku EQ.