With equal spine weights, retained spines gave more summed binocular calcium tuning change than turnover across 13 mice

We compared spine turnover with response changes in retained spines in a defined sum of visual responses. We used 138 segment intervals from 22 cells in 13 mice in released developmental mouse visual-cortex data. Each observed spine had one coefficient. Responses used source calcium noise units and one initial spine count denominator. We split total change exactly into retained response change and the difference between formed and lost responses. Separate repeat products estimated change energy under independent repeat errors. We shared the cross term equally between the two contributions. The retained-minus-turnover difference was 0.01476 source noise standard deviation squared, with a 95% whole-mouse bootstrap interval of 0.00445 to 0.02741. Mean initial retention was 64.3%. All distinct repeats gave a difference of 0.01417 [0.00399, 0.02675]. Source-accepted trials and both age intervals also gave positive differences. The primary turnover contribution was near zero and uncertain; its accepted-trial estimate was positive. Thus, turnover was substantial but gave less measured summed tuning change than retained responses under this readout. The result does not measure synaptic currents or establish a plasticity mechanism. Date-specific scales, shared preprocessing, partial fields and 13 mice limit inference. The package preserves code, source provenance, compact input, all scores, figures and repeat instructions.

structure and function · dendritic spines · visual cortex · development · spine turnover · calcium imaging · response stability · independent repeats