Despite the availability of an X-ray structure and many spectroscopic studies, important issues related to structural heterogeneity, excitonic structure, primary charge separation (CS), and excitation energy transfer dynamics of the isolated reaction center (RC) of photosystem II (PS II) remain unresolved. The issues addressed here include (1) whether the primary CS kinetics at low temperatures are highly dispersive (due to structural heterogeneity), as proposed by Prokhorenko and Holzwarth (J. Phys. Chem. B 2000, 104, 11563), and (2) the nature of the weak lowest-energy Q(y) absorption band at similar to684 nm that appears as a shoulder on the intense primary electron donor band (P680). Results of low-temperature nonphotochemical hole burning (NPHB) and triplet bottleneck hole burning (TBHB) spectroscopic experiments (including effects of pressure and external electric (Stark) fields) are presented for the RC from spinach with one of the two peripheral chlorophylls removed. Both NPHB and TBHB are observed with excitations within the P680 and 684 nm bands. Both types of hole spectra exhibit a weak dependence on the burn wavelength (lambda(B)) between 680 and 686 nm. Furthermore, the permanent dipole moment change (f Deltamu), as determined by Stark-NPHB spectroscopy, is identical (0.9 +/- 0.1 D) for the two bands, as are the linear electron-phonon coupling parameters (Huang-Rhys factors S-17 = 0.7 and S-80 = 0.2 for 17 and 80 cm(-1) phonons). These similarities, together with published fluorescence line narrowed spectra lead us to favor the gross heterogeneity model in which the 684 nm band is the primary electron donor band (P684) of a subset of RCs that may be more intact than P680-type RCs. It is concluded, based also on the linear pressure shift rates for the P680 and 684 nm bands, that population of either P680* (* equivalent to Q(y) state) or P684* results in both TBHB (due to charge recombination of the primary radical ion pair) and NPHB. It was found that the values of parameters (e.g., electron-phonon coupling, site distribution function) used to simulate the NPHB spectra also provided reasonable fits to the TBHB spectra. Acceptable theoretical simulations of the line-narrowed TBHB spectra were not possible using a single primary CS time. However, satisfactory fits (including lambda(B) and bum intensity dependences) were achieved using a distribution of CS times. The observed TBHB is due to P680- and P684-type RCs with the faster CS kinetics since the persistent nonphotochemical holes were saturated prior to measuring the TBHB spectra. (RCs exhibiting the most efficient NPHB have slower CS kinetics as well as higher fluorescence quantum yields.) For the TBHB spectra, the same distribution (Weibull) was used for the P680- and P684-type RCs. The distribution describes quite well the distribution of Prokhorenko and Holzwarth for CS times shorter than 25 ps. Finally, the data indicate that electron exchange contributes only weakly (relative to electrostatics) to the inter-pigment excitonic interactions.