Many different reasons have been proposed, but a generally regard one is the change in pH by increased amounts of cyanide.

Send more details on your process, cyanide concentrations, batch/continuous, DO, leach time, pro-ox etc. Maybe we can make more sense and even come up with some improvements.

I studied this extensively while working in Ecuador and do not believe that higher pH is the cause of the decelerating effect of gold solubility in higher concentrations of cyanide because we held the pH constant (at 10, at 10.5 and at 11 with lime and also with NaOH) in our studies. We found only that the solubility of copper increased dramatically with higher cyanide concentrations hence that may have been the cause.

I think, that first at all, we need to know, what is the operational context?, In this metallurgical process, there is so much copper, silver, mercury, and what about the oxygen concentration on the barren solution. The gold cyanidation, is a electrochemical process, so it´s very important to evaluate, with other electrochemical factors.

Interesting problem - probably passivation caused by the interaction between pyrite and cyanide (ferro-cyanides-iron hydroxides). High DO levels are also not good for gold dissolution in the presence of pyrite. Need to keep the CN and DO levels quite moderate.

Are you using lead nitrate to passivate the pyrite and what DO levels are you achieving?
Are you sure that it is pyrite and not marcasite or some pyrrhotite present?

These versions of iron sulphide can have a dramatic impact unless handled in specific ways (e.g. aeration prior to cyanide addition).

I think it's because of other secondary reactions that consume CN.