COS 116-9 - Estimating changes in above- and below-ground 3-D vegetation structure in a model savanna using ground penetrating radar and terrestrial laser scanning

Thursday, August 11, 2011: 4:20 PM
13, Austin Convention Center
Sean M. Thompson1, William T. Brademan2, Byron G. Barber2, Alfredo Delgado1, Michael A. Austin3, Dirk B. Hays4 and Robert A. Washington-Allen5, (1)Soil and Crop Sciences, Texas A&M University, College Station, TX, (2)Ecosystem Science and Management, Texas A&M University, College Station, TX, (3)Ecosystem Science and Management, Texas A & M University, College Station, TX, (4)Molecular and Environmental Plant Sciences, Texas A&M University, College Station, TX, (5)Geography, University of Tennessee, Knoxville, TN
Background/Question/Methods

Changes in the above ground structure and root morphology of savannas are unknown for wild and domesticated vegetation.  Previous work has shown that new tools including ground penetrating radar (GPR) and terrestrial laser scanning (TLS) are capable of non-invasively estimating biomass and mapping three-dimensional (3-D) above- and below-ground vegetation structure.  Consequently, our objective was to use a dual-channel (400 MHz and 1.5 GHz) GPR and a 55-KHz TLS to study changes in above- and below-ground 3-D structure, distribution, and biomass of two independent lines of wheat and clover and a line of potatoes as a model for dryland savannas of differing structural architecture. 

Results/Conclusions

A 10-m X 20-m plot was surveyed in a sandy savanna soil, divided into 6 rows with 4 beds per row for a total of 24 beds.  Eight beds per wheat line, 2 per clover line, and 4 for potatoes, were planted at 2 different seed densities and irrigated.  We expected that both sensors would detect a significant difference in above- and below-ground architecture of potatoes, clover, and wheat.  Two GPR transects per bed were scanned at each frequency for a total of 12 transects and the 10-m X 20-m plot was scanned along the 20-m length from the midpoint of the 10-m width with the TLS at a 30° vertical X 180° horizontal 30-m radius.  The GPR and TLS scans were repeated every two weeks to track phenological development.  Additionally, greenhouse beds were established, rhizotrons installed, and TLS and GPR (with the 1.5 GHz antenna) scanned at 2-week intervals.  GPR and TLS data were periodically processed to generate time series of 3-D visualizations of GPR amplitude returns, root mass, plant structure, and plant height.  Greenhouse plots were periodically harvested and excavated for above- and below-ground biomass calibration with GPR amplitude signal and TLS height measurements.  At harvest, plots were clipped and excavated and biomass determined.  We detected different root geometries in wheat and potatoes as expected and differences between clover and potatoes, but it was uncertain between clover and wheat.  We developed allometric equations for above- and below-ground estimation of biomass for each crop at different stages of phenology until harvest.  We also detected the evident differences between above ground architectures, because at TLS resolution (4.5-mm spot) species identification is possible.  Using TLS and GPR we were able to non-invasively estimate changes in above- and below-ground biomass in model savannas.

Copyright © . All rights reserved.
Banner photo by Flickr user greg westfall.