COS 76-2
Tree root ecology in multi-species agroecosystems: detection and biomass estimation with ground penetrating radar

Wednesday, August 12, 2015: 1:50 PM
301, Baltimore Convention Center
Kira A. Borden, Department of Geography, University of Toronto, Toronto, ON, Canada
Luke C.N. Anglaaere, Forestry Research Institute of Ghana, Kumasi, Ghana
Marney E. Isaac, Department of Physical and Environmental Science, University of Toronto-Scarborough, Toronto, ON, Canada

Perennial crops, such as the economically important cocoa tree (Theobroma cacao), dominant tropical landscapes. Yet, like many tropical tree species, we know very little about the root ecology of cocoa, specifically root structure and biomass. Generalized allometric or root:shoot relationships are conventionally used to account for the belowground components due to the challenging and destructive nature of measuring tree root systems in situ. However, empirical evidence that these relationships hold under managed agricultural scenarios is limited. In previous studies, we have shown the utility of the geo-imaging technology ground penetrating radar (GPR) for quantifying coarse root biomass of trees grown in agricultural landscapes. In this current study, we use GPR to estimate the coarse root biomass of productive cocoa trees in Ghana in order to 1) quantify root carbon storage and 2) assess if GPR can detect differential biomass allocation patterns in agroecosystems. To do this, GPR data were collected surrounding 15 individual cocoa trees grown i) in monoculture and ii) in mixture with shade trees (Terminalia ivorensis, Entandrophragma angolense). GPR signals reflect at coarse roots that contain more moisture than the soil and the signal response is used to calibrate biomass estimation across the study site.


Soil conditions and coarse root moisture content met necessary requirements for radar signal reflection and we found significant positive correlation between root biomass and GPR response. Our initial estimates show approximately 0.4 kg of carbon m-2 is stored in coarse root biomass and indicate differences in cocoa root biomass associated with different management. Biomass estimates will be compared to a subset of excavation measurements and evaluated against conventional allometric relationships. We will convert these area-based root biomass estimates to the individual plant scale via multiple spatial and analytical approaches based on empirical field data of tree size and neighboring trees.  Results from this study can inform carbon accounting of cocoa farming systems and indicate possible impacts on these carbon pools from differences in root biomass allocation under managed conditions.