The maximum rate for auditory nerve firing is northwards of 1000 spikes/sec, based on refractory effects (this will vary by species), but such a high rate is not sustainable because of adaptation.  With electrical stimulation, the degree of adaptation is reduced because one of its mechanisms, synaptic transmission between the hair cells and nerve, is effectively bypassed.  Even then short-duration max rates aren't much higher that 1 k sp/s.  Nevertheless, the properties of nerve firing are different at, say, 5000 pulses/sec than 1000 pulses/sec because of stochastic effects.  I'd recommend looking at a 2003 paper by Litvak and colleagues ("Desyhchronization of electrically...", JASA) and the models of Bruce (e.g "A stochastic model of the electrically stimulated auditory nerve...", IEEE Trans BME, 1999) for starters.  On a practical note, one potential benefit for high pulse rates appears to be a greater perceptual dynamic range, though this might come at the cost of temporal acuity.

See also this recent paper by Rosen on amplitude modulation detection at various carrier pulse rates: www.ncbi.nlm.nih.gov/pubmed/22367093.

Thank you so much for your help..

Based on my personal experience in recording single units, the maximum sustained firing rate in individual cochlear neurons (in mammalian species) will rarely exceed 200 spikes/s. There may be some pathological conditions (e.g. cochlear excitotoxicity) when refractory period limitations are minimal. Kiang reported modal values of inter-spike intervals between 4-7ms, equivalent to rates between 150-250 spikes/s.

Thank you Mr. Harrison;

That so intersting why we have so high firing rates for cochlear implants. Do you have any experince with this issue?

There are two major differences  between maximum firing rates in individual and NORMAL cochlear neurons and those issues related to  cochlear implant stimulation.

First there are no synapses involved in the latter. Second we are not talking abut activity generated in any INDIVIDUAL neuron, but the activation of thousands of spiral ganglion cells. Individual refractive period considerations are not so important.