Too late to edit this in, but a while back I came across this:
Chemical ecology of Cannabis


Ultraviolet radiation
Another stress to which plants are subject results from their daily exposure to sunlight. While necessary to sustain photosynthesis, natural light contains biologically destructive ultraviolet radiation. This selective pressure has apparently affected the evolution of certain defenses, among them, a chemical screening functionally analogous to the pigmentation of human skin. A preliminary investigation (Pate 1983) indicated that, in areas of high ultraviolet radiation exposure, the UV-B (280-315 nm) absorption properties of THC may have conferred an evolutionary advantage to Cannabis capable of greater production of this compound from biogenetic precursor CBD. The extent to which this production is also influenced by environmental UV-B induced stress has been experimentally determined by Lydon et al. (1987). Their experiments demonstrate that under conditions of high UV-B exposure, drug-type Cannabis produces significantly greater quantities of THC. They have also demonstrated the chemical lability of CBD upon exposure to UV-B (Lydon and Teramura 1987), in contrast to the stability of THC and CBC. However, studies by Brenneisen (1984) have shown only a minor difference in UV-B absorption between THC and CBD, and the absorptive properties of CBC proved considerably greater than either. Perhaps the relationship between the cannabinoids and UV-B is not so direct as first supposed. Two other explanations must now be considered. Even if CBD absorbs on par with THC, in areas of high ambient UV-B, the former compound may be more rapidly degraded. This could lower the availability of CBD present or render it the less energetically efficient compound to produce by the plant. Alternatively, the greater UV-B absorbency of CBC compared to THC and the relative stability of CBC compared to CBD might nominate this compound as the protective screening substance. The presence of large amounts of THC would then have to be explained as merely an accumulated storage compound at the end of the enzyme-mediated cannabinoid pathway. However, further work is required to resolve the fact that Lydon's (1985) experiments did not show a commensurate increase in CBC production with increased UV-B exposure.

This CBC pigmentation hypothesis would imply the development of an alternative to the accepted biochemical pathway from CBG to THC via CBD. Until 1973 (Turner and Hadley 1973), separation of CBD and CBC by gas chromatography was difficult to accomplish, so that many peaks identified as CBD in the preceding literature may in fact have been CBC. Indeed, it has been noted (De Faubert Maunder 1970) and corroborated by GC/MS (Turner and Hadley 1973) that some tropical drug strains of Cannabis do not contain any CBD at all, yet have an abundance of THC. This phenomenon has not been observed for northern temperate varieties of Cannabis. Absence of CBD has led some authors (De Faubert Maunder 1970, Turner and Hadley 1973) to speculate that another biogenetic route to THC is involved. Facts scattered through the literature do indeed indicate a possible alternative. Holley et al. (1975) have shown that Mississippi-grown plants contain a considerable content of CBC, often in excess of the CBD present. In some examples, either CBD or CBC was absent, but in no case were plants devoid of both. Their analysis of material grown in Mexico and Costa Rica served to accentuate this trend. Only one example actually grown in their respective countries revealed the presence of any CBD, although appreciable quantities of CBC were found. The reverse seemed true as well. Seed from Mexican material devoid of CBD was planted in Mississippi and produced plants containing CBD.

Could CBC be involved in an alternate biogenetic route to THC? Yagen and Mechoulam (1969) have synthesized THC (albeit in low yield) directly from CBC. The method used was similar to the acid catalyzed cyclization of CBD to THC (Gaoni and Mechoulam 1966). Reaction by-products included cannabicyclol, delta-8-THC and delta-4,8-iso-THC, all products which have been found in analyses of Cannabis (e.g., Novotny et al. 1976). Finally, radioisotope tracer studies (Shoyama et al. 1975) have uncovered the intriguing fact that radiolabeled CBG fed to a very low THC-producing strain of Cannabis is found as CBD, but when fed to high THC-producing plants, appeared only as CBC and THC. Labeled CBD fed to a Mexican example of these latter plants likewise appeared as THC. Unfortunately, radiolabeled CBC was not fed to their plants, apparently in the belief that CBC branched off the biogenetic pathway at CBD and dead ended. Their research indicated that incorporation of labeled CBG into CBD or CBC was age dependent. Vogelman et al. (1988) likewise report that the developmental stage of seedlings, as well as their exposure to light, affects the occurrence of CBG, CBC or THC in Mexican Cannabis. No CBD was reported.