Simulations of passive glutamate diffusion comparing reconstructed rod spherules with simplified models that represented the extracellular synaptic volume as a sphere with a narrow neck for an exit. (A) Illustration of reconstructed rod spherules 1 and 2. HC dendrites are in yellow and green, BP dendrites are in red and magenta, and ribbons are in blue. Below each spherule is the corresponding sphere model containing the same extracellular volume shown at the same scale. (B) Plot of the number of glutamate molecules that remained after the release of a vesicle in the synaptic cleft of rod 1 (D = 2 × 10−6 cm2/s, black line) or a sphere with the same extracellular volume (D = 8 × 10−6 cm2/s, blue line; D = 2 × 10−6 cm2/s, dashed red line). We simulated release in rod 1 at a site just beneath the center of the ribbon. (C) Graph of the glutamate decline in a sphere with the same extracellular volume as the invaginating synapse of rod 2 (D = 8 × 10−6 cm2/s, blue line; D = 2 × 10−6 cm2/s, dashed red line). This graph also shows the decline in glutamate following the release of a vesicle beneath the center of the ribbon in rod 2 with volume fractions of 0.11 (D = 2 × 10−6 cm2/s, black line), 0.2 (turquoise line), and 0.3 (purple line). Eliminating an absorptive perimeter that simulates avid Müller cell uptake had no effect on the time course of decay (gray trace). (D) Comparison of passive glutamate decline in the four rod spherules. Fitting the decays with single exponentials yielded the following time constants: rod 1: 10.64 ms; rod 2: 14.10 ms; rod 3: 7.45 ms; and rod 4: 8.89 ms. Each trace is an average of 12 simulations run using different seed values.