Biomass partitioning in Guineagrass: underlying mechanisms and prediction using the CROPGRO Perennial Forage Model

• Ano de defesa: 2020
• Autor(a) principal: Brunetti, Henrique Bauab
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Panicum maximum; Alongamento de colmo; Comprimento do pseudocolmo; Developmental stage; DSSAT system; Estágio de desenvolvimento; Gramíneas tropicais; Grassland; Partição de fotoassimilados; Pastagem; Photosynthate partitioning; Plataforma DSSAT; Pseudostem length; R/FR ratio; Razão V/VD; Stem elongation; Tropical grasses; Warm-season grasses
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/11/11139/tde-14082020-114741/

In pasture-based livestock systems, biomass partitioning affects the quality of forage offered to animals and the subsequent regrowth cycles. The objectives of this study were (i) to investigate the effects of the R/ FR ratio and of the developmental stage on stem elongation of guineagrass tillers, (ii) to investigate the effects of stem elongation on the tillers\' leaf appearance and elongation and (iii) to investigate and improve the ability of the CROPGRO Perennial Forage Model to simulate growth and biomass partitioning of guineagrass. For the objectives (i) and (ii), tiller-level data were collected in an irrigated Panicum maximum cv. Mombaça field evaluated during two vegetative cycles (summers 1 and 2) and one reproductive cycle (autumn). The R/FR ratio data was acquired with the aid of an Ocean Optics USB2000+ spectrometer. For each regrowth cycle, two to five times per week from 10 to 11h, five to ten measurements beneath the sward were recorded for each sample. For measurement of the true stem and pseudostem lengths, 10-25 representative tillers were weekly clipped and cut longitudinally. The developmental stage was visually observed with the aid of a Coleman XTB Stereoscope Microscope. Ten tillers/regrowth cycle were tagged for leaf appearance and elongation measurements. The variables were plotted against the accumulated growing degree-days (gdd) and models were fitted using the NLIN procedure of SAS 9.2. The true stem data were fitted in segmented (constant - linear increase) curves (adj-R2=0.95 and 0.99 for the summers 1 and 2, respectively). The elongation was not primarily related to R/FR ratio beneath the sward, but was satisfactorily explained by gdd (with a difference of only 35 gdd between the breakpoints of the summers 1 and 2). The true stem elongation did not shorten the pseudostem. During the reproductive cycle, the true stem elongation occurred concurrently with the appearance of flowers primordia (11.7h of daylength) and shortened the pseudostem. In the vegetative tillers, as opposed to the reproductive tillers, the true stem elongation did not shorten the final leaf length nor increased the leaf appearance rate. For the objective (iii), data from two experiments were used. One experiment was conducted with cultivar Tanzânia and the other with Mombaça, and they had contrasting harvest management and field conditions. Parameterization was done manually, targeting improvement in d-statistic and root mean square error (RMSE) averaged over both cultivars for aboveground biomass accumulation, leaf biomass, stem biomass, leaf area index (LAI), and leaf proportion on aboveground biomass. The major improvement in model performance was achieved by modifying the vegetative partitioning parameters between leaf and stem (increasing partitioning to leaf during early regrowth while increasing to stem during late regrowth). Productivity during cool, short daylength months was decreased by making photosynthesis and leaf area expansion more sensitive to lower temperatures and increasing the partitioning to storage organs during short days. Productivity under warm-season conditions was increased by decreasing leaf and stem senescence and reducing the nitrogen stress effect on leaf area expansion. The overall d-statistic averaged over the two cultivars increased from 0.86 to 0.93 and the RMSE values averaged over the two cultivars decreased from 2261 to 1768 kg ha-1 for aboveground biomass, from 1620 to 874 kg ha-1 for stem biomass, from 11.41 to 7.27% for leaf percentage, from 1.91 to 1.68 for LAI, but increased slightly from 1114 to 1172 kg ha-1 for leaf biomass. The improved model performance for both short and long harvest cycles will facilitate further use for evaluating various management strategies for these grasses.