MEASURING HETEROTROPH-INDUCED SOURCE-SINK RELATIONSHIPS IN PANICUM-COLORATUM WITH C-11 TECHNOLOGY

We report a synthesis from three series of experiments on source-sink relationships in Panicum coloratum L., a C-4 tropical grass obtained f rom the Serengeti grasslands of Africa. Studies on C-11 real-time anal yses of P. coloratum to determine above-ground effects of grasshopper grazing and belowground effects of mycorrhizal inoculation and nematod e feeding provided the database. A series of multi- and univariate sta tistical investigations of all available experimental data described r esponses of leaves, stems, and roots to these biological stresses. Fro m a principal components analysis we have shown differences in distrib ution of C source-sink locations along three principal component axes, which accounted for 84% of the experimental variance. The first and s econd components (62% of variance) described C allocation to leaf, ste m, and root sinks. The third component (22% of variance) showed a meta bolic dichotomy between leaf starch sinks and labile carbon pools thro ughout the plant. We use the three principal components from a C-11 th ree-compartment model describing leaf, stem, and root C source and sin k variables to present patterns, or fingerprints, of responses to the experiments. Time of day, treatment class, number of days since transp lanting, and ecotype controlled a large amount of the overall variatio n in plant C fixation and reallocation. A comparison of C-12 leaf carb on exchange rates (CER) measured with an infrared gas analyzer and C-1 1 rates showed a high positive correlation. Slopes for grasshopper gra zing, mycorrhizal inoculation experiments, and nematode feeding showed almost identical results; however, differences in the intercept devel oped as a function of ecotype. We noted a significantly lower intercep t in morning studies, but no differences in the slope for morning comp ared to afternoon studies. CER and all C-11 variables for grasshopper and nematode experiments showed a lower coefficient of variation (CV) than controls; mycorrhizae had slightly lower CV values for C-11 varia bles and higher for CER. We conclude that C-11 experiments provide the base for developing laboratory, field, and modeling studies to incorp orate aggregations of real-time C transfers within plants responding t o biological stresses, including those of heterotrophs.