RESUMEN
How trees sense source-sink carbon balance remains unclear. One potential mechanism is a feedback from non-structural carbohydrates regulating photosynthesis and removing excess as waste respiration when the balance of photosynthesis against growth and metabolic activity changes. We tested this carbohydrate regulation of photosynthesis and respiration using branch girdling in four tree species in a wet tropical rainforest in Costa Rica. Because girdling severs phloem to stop carbohydrate export while leaving xylem intact to allow photosynthesis, we expected carbohydrates to accumulate in leaves to simulate a carbon imbalance. We varied girdling intensity by removing phloem in increments of one-quarter of the circumference (zero, one--quarter, half, three-quarters, full) and surrounded a target branch with fully girdled ones to create a gradient in leaf carbohydrate content. Light saturated photosynthesis rate was measured in situ, and foliar respiration rate and leaf carbohydrate content were measured after destructive harvest at the end of the treatment. Girdling intensity created no consistent or strong responses in leaf carbohydrates. Glucose and fructose slightly increased in all species by 3.4% per one-quarter girdle, total carbon content and leaf mass per area increased only in one species by 5.4 and 5.5% per one-quarter girdle, and starch did not change. Only full girdling lowered photosynthesis in three of four species by 59-69%, but the decrease in photosynthesis was unrelated to the increase in glucose and fructose content. Girdling did not affect respiration. The results suggest that leaf carbohydrate content remains relatively constant under carbon imbalance, and any changes are unlikely to regulate photosynthesis or respiration. Because girdling also stops the export of hormones and reactive oxygen species, girdling may induce physiological changes unrelated to carbohydrate accumulation and may not be an effective method to study carbohydrate feedback in leaves. In three species, removal of three-quarters of phloem area did not cause leaf carbohydrates to accumulate nor did it change photosynthesis or respiration, suggesting that phloem transport is flexible and transport rate per unit phloem can rapidly increase under an increase in carbohydrate supply relative to phloem area. Leaf carbohydrate content thus may be decoupled from whole plant carbon balance by phloem transport in some species, and carbohydrate regulation of photosynthesis and respiration may not be as common in trees as previous girdling studies suggest. Further studies in carbohydrate regulation should avoid using girdling as girdling can decrease photosynthesis through unintended means without the tested mechanisms of accumulating leaf carbohydrates.