Potential photochemical production rates of total (NO-scavengeable) free radicals were surveyed underway (>900 points) in the eastern Caribbean and Orinoco delta in spring and fall 1988. These data document seasonal trends and large-scale (~10--100 km) variability in the pools of sunlight-generated reactive transients, which probably mediate a major portion of marine photoredox transformations. Radical production potential was detectable in all waters and was reasonably quantifiable at rates above 0.25 nmol L-1 min-1 sun-1. Radical production rates varied from ~0.1--0.5 nmol L-1 min-1 of full-sun illumination in ''blue water'' to>60 nmol L-1 min-1 in some estuarine waters in the high-flow season. Qualitatively, spatiotemporal potential rate distributions strikingly resembled that of ''chlorophyll'' (a riverine-influence tracer of uncertain specificity) in 1979--1981 CZCS images of the region <M¿ller-Karger et al., 1988> at all scales. Basin-scale occurrence of greatly enhanced rates in fall compared to spring is attributed to terrestrial chromophore inputs, primarily from the Orinoco River, any contributions from Amazon water and nutrient-stimulus effects could not be resolved. A major part of the functionally photoreactive colored organic matter (COM) involved in radical formation clearly mixes without massive loss out into high-salinity waters, although humic acids may flocculate in estuaries. A similar conclusion applies over smaller scales for COM as measured optically <Blough et al., this issue>.

Furthermore, optical absorption and radical production rates were positively correlated in the estuarine region in fall. These cruises demonstrated that photochemical techniques are now adequate to treat terrestrial photochemical chromophore inputs as an estuarine mixing problem on a large scale, though the ancillary data base does not currently support such an analysis in this region. Eastern Caribbean waters are not markedly more reactive at comparable salinities than waters of the Gulf of Maine and North Atlantic Bight, despite large inputs of colored waters from two large tropical rivers with substantial ''black water'' tributaries. Other sources of reactive COM, such as grazing, sedimentary diagenesis, and ''marine humus'' may increase temperate waters' photoreactivity; alternatively, northern waters may be chromophore-rich because they are light-poor and photobleaching is a major sink of photoreactive COM. ¿ American Geophysical Union 1993