As above and belowground resource availability vary in space and time in most ecosystems, I investigated the response of vesicular-arbuscular mycorrhizas (VAM) to variations in the belowground resource phosphorus (P) and two aboveground resources, light and atmospheric carbon dioxide concentration (CO2) in both single species and whole plant communities and in field and controlled environment conditions. The tropical wet forest tree Inga coruscans responds to variation in plant light environment with changes in the development of VAM fungal structures and in VAM root colonization. VAM lengths were greater in light gap than in understory saplings and VAM length was inversely correlated with canopy cover. Light gap saplings had significantly more intercellular fungal coils in their roots than did understory saplings. Among shadehouse-grown seedlings, VAM colonization increased with light intensity, from a mean of 10.5 cm colonized root at 9.05 $mu$mol photons m$sp^-2~$s$sp^-1~$ to 211.3 cm at 410 $mu$mol photons m$sp^-2~$s$sp^-1~.$ Addressing mycorrhizal response to belowground resource distribution I studied VAM ecology in a serpentine annual grassland. VAM spores germinate with the first rains: two weeks into the wet season densities dropped from 110 spores g$sp^-1~$ soil at the end of the dry season to under 80 spores g$sp^-1~$ and plant roots were already 20 to 30 percent colonized by VAM fungi. VAM colonization increased to 70 percent of total root length at peak plant flowering. Spore abundance was negatively correlated with both extractable soil nitrate and phosphate, and was low on harvester ant mounds. The spatial patterning of these ant mounds is similar to patterns of nutrient availability and VAM spore density, thus ant activity may shape VAM ecology in this ecosystem. Finally, I examined VAM response to crossed gradients of plant-fixed carbon and soil-available P. In controlled environments, when pCO$sb2$ and plant carbon acquisition were ambient plants direct carbon resources to VAM when P is low and to roots when P is available, while elevated CO2 plants allocate to roots when P is low and maintain VAM regardless of P treatment. By contrast, studies in two grassland ecosystems with differing nutrient availability revealed no significant differences in VAM colonization between either pCO$sb2$ treatments or grasslands.