Influence of Irrigation and Fertilization on the Belowground Carbon Allocation in a Pine Plantation

The aboveground and belowground productivity of forest systems are interlinked through complex feedback loops involving tree, soil and environmental factors. With a predicted significant change in environmental conditions through the enhanced greenhouse effect, it is important to understand the response of forest systems to these new conditions. An increase in atmospheric CO2 is predicted to increase photosynthesis, and therefore whole plant productivity at the individual tree level. However this increase in photosynthesis may result in greater requirements for nutrients, particularly nitrogen (N). In order to acquire any additional available N, trees may respond by increasing their proportional allocation of C belowground to the root system. This study aimed to quantify the belowground C allocation in a mature forest system consisting of a single species on a single site, but with different levels of water and nutrient stress. The belowground carbon dynamics of a range of irrigated and fertilized Pinus radiata stands in Australia were investigated during 1992 and 1993. Belowground carbon allocation was estimated using the model proposed by Raich and Nadelhoffer (1989) where belowground C allocation is the difference between soil respiration and carbon input through litterfall, plus coarse root production and an adjustment for any change in soil and litter layer carbon pools. This model is best described by the equation: Belowground C = Csoilresp – Clitterfall + Ccoarseroot+ ∆Cforest floor+ ∆Csoil Soil respiration, measured using a modified soda lime absorption method either every 2 weeks or every 4 weeks for 2 years, showed a range in daily soil C flux from 137 – 785 mgCO2.m-2.h-1. Soil respiration showed seasonal trends with summer highs and winter lows. Limited fine root biomass data could not indicate a strong relationship between measured soil respiration and fine root (&gt2mm diameter) biomass. Fifty three percent of the variation in soil respiration measurements in irrigated treatments was explained by a linear relationship between soil respiration, and soil temperature at 0.10 m depth and litter moisture content. In non-irrigated treatments, 61% of the variation in soil xixrespiration measurements was explained by a linear relationship between soil temperature at 1 cm depth and soil moisture content. Inter-year variation was considerable with annual soil respiration approximately 20% lower in 1993 compared with 1992. Annual soil C flux was calculated by linear interpolation and ranged from 3.4 – 11.2 tC ha-1 across the treatments. Soil C pools remained unchanged over 10 years between 1983 and 1993 for all combinations of irrigated and fertilized stands, despite significant aboveground productivity differences over the decade. Measurements of standing litter showed a change between 1991 and 1993 for only 2 out of the 10 treatments. These two treatments had belowground C allocation estimated both with and without an adjustment for a change in standing litter. Annual litterfall C ranged almost four fold from 0.6 – 2.2 tC ha-1 between the treatments in 1992 and 1993, and fell within the ranges of measured litterfall over 10 years at the field site. Again inter-year variation was large, with the 1993 litterfall values being approximately 97% greater across all treatments compared with 1992 values. Belowground carbon allocation was calculated using C fluxes measured at the field site, and ranged 3 fold from 4.4 – 12.9 tC ha-1 between the treatments during 1992 and 1993. In 1993 the belowground C allocation was approximately 30% lower across all treatments compared with 1992 calculations. This was due to an approximate 23% reduction in annual soil C flux, a 97% increase in litterfall C and an 18% reduction in coarse root production between 1992 and 1993. The field site was N limited, and differences in belowground C allocation could be shown across irrigated treatments with different N limitations. As N availability increased belowground C allocation was decreased in the irrigated treatments. It was difficult to determine differences in belowground C allocation caused by water stress as the effects of water and N limitation were confounded. An increase in N availability generally indicated an increase in coarse root and litterfall C production, which were reflected in increased aboveground productivity. In high N treatments the coarse root fraction of belowground C allocation comprised approximately 50% of the total belowground C allocation, whereas in the N stressed treatments coarse roots only comprised 20% of the total belowground allocation The mechanistic model BIOMASS was used to estimate annual gross primary productivity (GPP) for the different treatments at the field site. BIOMASS estimated GPPs of between 30-38 tC ha-1 for the different treatments during 1992 and 1993. The measured belowground carbon allocation ranged from 16 – 40 % of simulated GPP, with the lower proportion allocated belowground in the irrigated and high fertility stands. Aboveground competition through the absence of thinning also appeared to reduce allocation belowground in non-irrigated stands. A direct trade off between bole and belowground C could not be demonstrated, unless data were separated by year and by the presence or absence of irrigation. Where data were separated in this manner, only three data points defined the reasonably strong, negative relationship between bole and belowground C. The value of this relationship is highly questionable and should be interpreted with caution. Thus a decrease in belowground C allocation may not necessarily indicate a concomitant increase in bole C allocation. Inter-year variation in a number of C pools and fluxes measured at the field site was at least as great as the variation between stands having different water and N limitation. Extrapolation of belowground productivity estimates from a single years data should be undertaken cautiously. The work undertaken in this study indicated that for a given forest stand in a given soil type, an increase in N availability reduced the absolute and relative C allocated belowground. However this decrease in C belowground may not directly translate as an increase in stem growth or increased timber production. Forest productivity in an enhanced greenhouse environment is likely to result in an increased allocation of C belowground due to increased N limitation, unless adequate N is present to support a more active canopy. Further work is required to more fully understand the dynamics of the belowground system in a changing environment. However further research should focus on mature forest systems in order to isolate the impacts of natural ageing changes from perturbation effects on the forest system. This would be best undertaken in long term monitoring sites where a C history of the stand may be available