Root restriction and root-shoot relationships in tomato (Lycopersicon esculentum Mill.) : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Horticultural Science at Massey University

Abstract

The potential for controlling plant growth and productivity by manipulating root growth and development has not been realised because of a lack of understanding of how root growth influences shoot growth. Until such responses are understood, matching container design and volume to desired plant output will continue to be based solely on anecdotal evidence. A series of experiments were conducted to explore the role of physical root restriction on the vegetative growth and development of tomato (Lycopersicon esculentum Mill. 'Moneymaker'). Concurrent with these experiments, a statistical model was developed for non-destructively estimating leaf area, cluster analysis was adapted to improve experimental precision, and an improved form of growth analysis developed. Additionally, a review of oxygen and major nutrient uptake rates by tomato established the operational parameters of a hydroponic system developed specifically for the study. Rooted tomato cuttings were grown in 0.025 or 10 litre (control) containers in the hydroponic system. After 31 days in 0.025 litre containers, plants were de-restricted into either 0.05 or 10 litre containers, or retained in the 0.025 litre containers. Plants with physically restricted root systems had lower total plant biomass and total leaf area, were shorter in both height and total root length, and had fewer roots, leaves, and lateral shoots than unrestricted plants. Restriction reduced root number after 31 days, but reductions in root length and dry biomass did not occur until after 45 days. Leaf dry biomass was reduced in restricted plants after 45 days; reductions in stem height, leaf area, number and total dry biomass) were apparent after 67 days. Short periods (31 days) of root restriction had long term (67-99 days) effects on leaf growth. Leaf expansion was more sensitive than leaf biomass accumulation to root restriction. A strong linear relationship, independent of root restriction, was observed between the relative rates of root elongation and leaf expansion. Similar relationships with the relative rates of increase in root number and dry biomass were due to their covariance with root elongation. These data are consistent with the hypothesis that root elongation is functionally linked to leaf expansion via the synthesis of hormones in actively growing root apices. The influence of partial root restriction on leaf expansion was also examined. One or both halves of a split root system was enclosed in a 30 cm3 polyethylene cell. Leaf expansion was reduced in plants with only a portion of their total root system physically restricted. Compensatory growth in the unrestricted portion of the root systems resulted in total root growth at final harvest being similar to plants with all their root system unrestricted. Analysis of the relative rate of leaf expansion (RA) of individual leaves along the stem axis revealed two distinct phases in response to root restriction. In the first phase, apparent about 28 days after treatments were initiated (DAI) and observed in leaves that started expansion 3, 7, and 14 DAI, RA was reduced in plants with one or both root sub-systems in a restriction cell. The second phase, detected 42 DAI and observed in leaves that started expanding 21 and 28 DAI, was characterised by a higher RA in plants with a portion of their root system restricted compared to unrestricted plants. Proportionately more assimilate was partitioned to stems of plants with two restricted root sub-systems compared to plants with either a single or non-restricted root sub-system. No differences in leaf water potential or photosynthesis of leaves were observed among treatments. Conclusions drawn from these data support the involvement of chemical signals in maintaining coordination between root and shoot growth in container-grown plants. These conclusions are discussed with reference to the literature, and a model is proposed to explain root-shoot coordination in terms of root-sourced cytokinin and shoot-sourced auxin. Avenues for future research to test hypotheses arising from this model are identified and discussed, as are possible horticultural ramifications. Emphasis was placed in the study on improving analytical methodology of growth analysis of whole-plant studies. Experimental precision was increased in these experiments by using cluster analysis to allocate plants to blocks based on leaf area, with a developmental study showing that the mean coefficient of variation of groups formed from cluster analysis was between two and fives times smaller than that of groups formed from visual assessment. A statistical model for non-destructively estimating the leaf area of tomatoes was developed based on the length of the mib-rib of each compound leaf and its position on the stem. Although the model was accurate to within about 2.5% of actual leaf area, it was not stable in time. It was concluded that when non-destructive estimation of tomato leaf area is required, the prediction model must be developed while the main experiment is being conducted. A hybrid method of growth analysis, incorporating both functional and univariate statistical approaches, provided more flexibility and information than standard functional or classical analytical methods. The hybrid method yielded replicated estimates of growth analysis indices, providing opportunity for further evaluation of the derived data using multivariate analytical techniques including path, canonical correlation, and canonical discriminant analysis. keywords: allometric relationships, assimilate partitioning, biometrics. Chanter function, cluster analysis, containerised plants, hydroponics, leaf expansion, local error control, plant growth analysis, relative growth rate, Richards function.