Retinoic acid treatment results in significant differences in the two-dimensional pattern of photoreceptors. (A) Conformity Ratio (CR) analysis of red, blue, and UV cones at 60 hpf. RA treatment caused significantly reduced regularity in spacing. Theoretical fate changes of 25% and 50% (see Methods) also resulted in significantly reduced CRs. (B) Nearest Neighbor Distance analysis (NND) of rods in singly-hybridized retinas at 60 hpf. RA treatment resulted in a significant decrease in NNDs (Additional File 1, Table S1; p < 0.001). (B') Autocorrelative NND analysis of rod patterns from doubly-hybridized retinas (rod and red cone opsin). Theoretical fate changes (25%, 50% of red cones to rods) also resulted in significantly reduced mean NNDs. (C) Density Recovery Profile analysis (DRP) of rods in singly hybridized retinas at 60 hpf. RA caused a significantly smaller Effective Radius (Reff; Additional File 1, Table S1; p < 0.05); this change does not take place in untreated retinas sampled later, at 75 hpf (C') Autocorrelative DRP analysis for rod patterns from doubly-hybridized retinas (rod and red cone opsin). Theoretical fate changes also resulted in significantly reduced mean Reffs. (D) Quadrat analysis compares long-range pattern of rods and cones in control (solid lines) and RA-treated (dotted lines) embryos. With RA, long-range pattern of rods becomes more regular while those of all cone types become more random. (E) Quadrat analysis of theoretical red cone patterns reveals decreasing pattern regularity (compare red line to green) as a consequence of an increasing percentage of loss from the red cone mosaic.