The effect of elevated CO2 and N availability on tissue concentrations and whole plant pools of carbon-based secondary compounds in loblolly pine (Pinus taeda)

We examined the extent to which carbon investment into secondary compounds in loblolly pine (Pinus taeda L.) is changed by the interactive effect of elevated CO2 and N availability and whether differences among treatments are the result of size-dependent changes. Seedlings were grown for 138 days at two CO2 partial pressures (35 and 70 Pa CO2) and four N solution concentrations (0.5, 1.5, 3.5, and 6.5 mmol l-1 NO3NH4) and concentrations of total phenolics and condensed tannins were determined four times during plant development in primary and fascicular needles, stems and lateral and tap roots. Concentrations of total phenolics in lateral roots and condensed tannins in tap roots were relatively high regardless of treatment. In the smallest seedlings secondary compound concentrations were relatively high and decreased in the initial growth phase. Thereafter condensed tannins accumulated strongly during plant maturation in all plant parts except in lateral roots, where concentrations did not change. Concentrations of total phenolics continued to decrease in lateral roots while they remained constant in all other plant parts. At the final harvest plants grown at elevated CO2 or low N availability showed increased concentrations of condensed tannins in aboveground parts. The CO2 effect, however, disappeared when size differences were adjusted for, indicating that CO2 only indirectly affected concentrations of condensed tannins through accelerating growth. Concentrations of total phenolics increased directly in response to low N availability and elevated CO2 in primary and fascicular needles and in lateral roots, which is consistent with predictions of the carbon-nutrient balance (CNB) hypothesis. The CNB hypothesis is also supported by the strong positive correlations between soluble sugar and total phenolics and between starch and condensed tannins. The results suggest that predictions of the CNB hypothesis could be improved if developmentally induced changes of secondary compounds were included.