The Effect of Canopy Architecture and Seasonal Variations on Several Seed Quality Attributes in Soybean (Glycine Max L. Merr.)
Gutema, Erenso Degu (2006) The Effect of Canopy Architecture and Seasonal Variations on Several Seed Quality Attributes in Soybean (Glycine Max L. Merr.). PhD thesis, Universiti Putra Malaysia.
Producing high quality soybean seed in the hot humid tropics is no easy task. During seed production, several environmental factors and plant morphological characteristics can exert their influences on seed quality. A study was undertaken at Universiti Putra Malaysia to study the effect of canopy architecture and seasonal variations on several seed quality attributes in soybean (GZycine max L. Merr.). Four soybean cultivars namely, Palmetto, AGS190, Deing and Cikurai were grown in the field for four seasons during 2003 and 2004. Four levels of defoliation treatments [0% defoliation (*170.89pmol m'2s" light intensity), 25% defoliation (*324.33pmol m'2s-' light intensity), 50% defoliation (*473.01pmol m-2s-' light intensity) and 75% defoliation (+642.84pmol m'2s" light intensity], were iniposed at the pod initiation stage (R3). Weather factors such as light intensity, canopy, air and soil temperatures, canopy and air relative humidity, soil moisture and leaf area index were recorded at seven-day intervals starting from the imposition of defoliation until plants reached physiological maturity (R7). Seeds harvested at harvest maturity (R8) were used to determine seed yield, viability, vigour, 1 00-seed weight and for Phomopsis bioassay. Phomopsis sp. seed infection was predicted using weather factors and leaf area index. Scanning electron microscopy (SEM) was used to study the progression and colonization of Phomopsis sp. on the stem, pod and seed starting from R3 until R8. Defoliation treatments were found to have inconsistent effects on seed yield (kg ha-') and pods per plant for AGS190, Deing and Palmetto. Pod number of Palmetto was affected during season Ill only whereas the pod number for Cikurai was significantly affected for all seasons. However, defoliation treatments affected 100-seed weight for all cultivars except for AGS 190 (season I), Deing (season IV) and Cikurai (season 111). From the combined analysis of data over the four seasons, defoliation improved percent seed germination from 6.8 to 13.2%. Increasing the level of defoliation resulted in increasing percent germination and 3-day seedling height and reduced Phomopsis incidence for all cultivars. The highest germination was recorded during season I1 which coincided with the least level of Phomopsis sp. seed infection. Moreover, this disease was influenced by seasonal variations. The disease incidence was high during seasons Ill (51.3%) and IV (49.5%) characterized by high rainfall during seed development and maturation as compared to seasons I and 11 which encountered low rainfall situations (33.3% and 32.5%, respectively). AGSl90, a large seeded cultivar, was severely affected by the seed-borne disease (51.5% infection) whereas Deing, a small seeded cultivar, was the least affected (34.7%). Defoliation treatments increased light intensity within the plant canopy for all cultivars studied. Light intensity and canopy temperature revealed negative correlation with percentage Phomopsis incidence indicating that high light intensity and temperature inside plant canopy reduced Phomopsis sp. seed infection. On the contrary, positive relationship was observed between canopy and air relative humidity, soil moisture and leaf area index with percentage Phomopsis incidence. From the stepwise multiple regression analysis, Phomopsis sp. seed infection can be predicted by leaf area index, soil moisture and canopy relative humidity; while its reduction can be predicted by increased light intensity and canopy temperature during seed development and maturation. Scanning electron microscopy revealed that hngi progression and colonization started at different growth stages for different plant parts. Stems were infected during the early reproductive stage (R3) whereas pods became noticeably infected during the full seed stage (R6) and seeds were the last to be infected. The hngi were internally- borne within the infected seeds. Fungal hyphae were observed externally on the surface of the plant parts studied and internally both in the pod and in all the three layers of the seed coat: palisade cell, hourglass cell and parenchyma cell layers. Fungi colonization was highest at the late plant growth stages of R7 and R8. A progressive increase in pod and seed infection was detected during subsequent growth stages between R6 and R8. From the pod, the pathogen can infect and colonize the seed. The SEM results suggested that the reproductive growth period of R6 to R7 was more critical with respect to Phomopsis sp. seed infection than earlier reproductive growth periods, since more severe colonization of pods and seeds took place at the later stages of plant growth. SEM revealed that stem infection allowed buildup of inoculum for subsequent infection of the pod, whereas pod infection was necessary for further infection of seeds. Since pod infection is the prerequisite for seed infection, this study suggests that fbngicide would be best applied between R4 and R5 before the seed-borne fungi reach the seed during R6 and the subsequent growth stages. Prediction model based on four seasons' data accurately described the relationship between the environmental conditions and leaf area index during seed development and maturation and the levels of seed infection by Phomopsis sp. Moreover, the model fitted well with the field and laboratory data collected. However, this model needs to be tested at multilocational trials for validity. The results of the present study have shown that plant canopy modification through defoliation appears to improve quality of seeds produced under wet and warm tropical environments. In addition, the study also suggested that growing of soybean cultivars with open canopies and having low leaf area index, coupled with rain-free harvesting seasons can result in the production of high quality seeds. Although the prediction model so developed in this study needs to be tested for validity at different locations and variable environments, it has the potential to be used as a practical tool in plant disease forecasting programs.
Repository Staff Only: Edit item detail