(a) Left: Firing rate maps in A|B and AB. Cell number and rat identity are indicated at the top (t = tetrode, c = cell). Color coded from dark blue (0) to dark red (peak firing rate). Peak Firing rates are indicated at the top and bottom right. Scale bar, 50 cm. Right: Sliding correlation heat map for the cell to the left color-coded from dark blue to dark red (color bar). Scale bar, 50 cm. (b) Distribution of values for mean grid spacing in A, B, and AB across the entire cell population. A wide range of values, with no particular bias, is represented. (c) Ratemaps in A and B were in most cases represented by non-identical grid patterns, although the correlation between rate maps for A and B (shown in c) was higher than expected by chance (Pearson product-moment correlation A vs. B: 0.237 ± 0.023, one-sample two-sided Student’s t-test: t(127) = 10.4, P = 1.14 x 10−18; correlation A vs. 180-degree rotation of B: 0.002 ± 0.021; two-sample two-sided Student’s t-test for correlation A vs. B against correlation A vs. B rotated: t(254) = 7.6, P = 7.5 x 10−13). (d) Line plot showing change in mean correlation between A|B and AB in the central 10% bands and the distal 10% bands as a function of grid spacing. Grid spacing was sorted into bins of 10cm width (mean ± s.e.m., n = 128 cells, 10 rats). Note that, for cells with grid spacing < 90–100 cm, central correlations were lower than distal correlations across the entire range of values for grid spacing, suggesting that reorganization is stronger in the centre than the periphery across several modules. For cells with larger values for grid spacing, correlation values may be inaccurate since larger grids have fewer fields, which mostly touch the distal walls. This, and the frequent absence of fields in the center, may cause inaccuracy in estimates of central field translation.