Purposeful release of CO2 into the ocean may occur in attempts to sequester it and mitigate global climate change, whereas accidental release might occur if underground CO2 used in enhanced oil recovery or geological sequestration were to escape from reservoirs on the ocean floor. Concentrated, continuous release from either scenario would result in a rising plume of CO2 droplets or bubbles, which dissolve and may react to form solid CO2-hydrates. At various points in its ascent, CO2-enriched seawater with hydrate particles peels away from the plume, forming lateral intrusions which disperse the CO2 into the surrounding ocean. The present work investigates, both theoretically and experimentally, regimes of CO2 dispersal for real-world ocean conditions. Our theoretical results map out CO2 dispersal regimes for release conditions such as depth, volume flux, and droplet or bubble size. We present a new regime in which dissolution or reaction, as opposed to ambient density stratification, dominates CO2 dispersal. We find that in deep ocean conditions these plume structures collapse, producing a sinking current of heavier CO2-laden seawater and droplets. Although some droplets will escape and rise from this current, most will continue to dissolve and be transported to large depth by the heavy, sinking flow. This flow increases the effectiveness of CO2 transport to and retention at depth in the ocean. Experimental work investigates the new regime where CO2 dissolution and reaction dominate large-scale CO2 dispersal. We use a multi-phase system consisting of calcium carbonate particles, CO2 bubbles, and an acidic aqueous solution to mimic this regime for laboratory conditions at atmospheric pressure. Size-specific calcium carbonate particles are released from the top of a tank, forming a descending plume whose fluid dynamics are similar to those of deep ocean plumes. The descending particles react to form CO2 bubbles, which oppose the buoyancy of our plume similar to how CO2-hydrates oppose the buoyancy of a deep ocean CO2 plume. We use digital image processing to investigate the time-averaged behaviour of such plumes, and draw implications for the modeling of CO2 sequestration.