The intriguing difference between far-infrared photoconductivity spectroscopy and absorption spectroscopy
in the measurement of the magnetoplasmon frequency in GaAs quantum wells reported by
Holland et al. [Phys. Rev. Lett. 93, 186804 (2004)] remains unexplained to date. This Letter provides a
consistent mechanism to solve this puzzle. The mechanism is based on the electron reservoir model for the
integer quantum Hall effect in graphene [Phys. Lett. A 376, 616 (2012)]. We predict sharp kinks to appear
in the magnetic induction dependence of the magnetoplasmon frequency at very low temperatures such as
14 mK in the same GaAs quantum well sample used by Holland et al..