Individual plant cells have a genetic circuit, the circadian clock, that times key processes to the day-night cycle. These clocks are aligned to the day-night cycle by multiple environmental signals that vary across the plant. How does the plant integrate clock rhythms, both within and between organs, to ensure coordinated timing? To address this question, we examined the clock at the sub-tissue level across Arabidopsis thaliana seedlings under multiple environmental conditions and genetic backgrounds. Our results show that the clock runs at different speeds (periods) in each organ, which causes the clock to peak at different times across the plant in both constant environmental conditions and light-dark (LD) cycles. Closer examination reveals that spatial waves of clock gene expression propagate both within and between organs. Using a combination of modeling and experiment, we reveal that these spatial waves are the result of the period differences between organs and local coupling, rather than long-distance signaling. With further experiments we show that the endogenous period differences, and thus the spatial waves, can be generated by the organ specificity of inputs into the clock. We demonstrate this by modulating periods using light and metabolic signals, as well as with genetic perturbations. Our results reveal that plant clocks can be set locally by organ-specific inputs but coordinated globally via spatial waves of clock gene expression.PLOS Biology | https://doi.org/10.days at near-cellular resolution (Materials and methods). This reporter line was chosen because of its strong expression level and its similar spatial expression to other clock components [5].In order to observe the endogenous component of the rhythms, we first imaged seedlings under LL, having previously grown them under LD cycles (LD-to-LL; Fig 2A and Materials and methods). Under the LD-to-LL condition we observed phase differences of GI::LUC expression between organs ( Fig 2B and 2C). The cotyledon and hypocotyl peaked before the root, but the tip of the root peaked before the middle region of the root (Fig 2C, S1 Fig, and S1 Video). Furthermore, we observed a decrease in coherence between regions over time, with a range between the earliest and latest peaking region of 4.92 ± 3.79 h (mean ± standard deviation) in the first and 18.36 ± 5.67 h in the final oscillation. This is due to the emergence of period differences between all regions ( Fig 2D). The cotyledon maintained a mean period of 23.82 ± 0.60 h, whereas the hypocotyl and root ran at 25.41 ± 0.91 h and 28.04 ± 0.86 h, respectively. However, the root tip ran slightly faster than the middle of the root, with a mean period of 26.90 ± 0.45 h, demonstrating the presence of endogenous period differences across all regions. We verified that our results were not specific to the GI::LUC reporter, as we observed similar differences in periods and phases across the plant using luciferase reporters for promoter activity of the core clock genes PSEUDO-RESPONSE REGULATOR 9 (PRR9) [31], TI...