Using High-Throughput Phenotyping to Explore Growth Responses to Mycorrhizal Fungi and Zinc in Three Plant Species

Abstract

<jats:p>There are many reported benefits to plants of arbuscular mycorrhizal fungi (AMF), including positive plant biomass responses; however, AMF can also induce biomass depressions in plants, and this response receives little attention in the literature. High-throughput phenotyping (HTP) technology permits repeated measures of an individual plant’s aboveground biomass. We examined the effect on AMF inoculation on the shoot biomass of three contrasting plant species: a vegetable crop (tomato), a cereal crop (barley), and a pasture legume (Medicago). We also considered the interaction of mycorrhizal growth responses with plant-available soil zinc (Zn) and phosphorus (P) concentrations. The appearance of a depression in shoot biomass due to inoculation with AMF occurred at different times for each plant species; depressions appeared earliest in tomato, then Medicago, and then barley. The usually positive-responding Medicago plants were not responsive at the high level of soil available P used. Mycorrhizal growth responsiveness in all three species was also highly interactive with soil Zn supply; tomato growth responded negatively to AMF inoculation in all soil Zn treatments except the toxic soil Zn treatment, where it responded positively. Our results illustrate how context-dependent mycorrhizal growth responses are and the value of HTP approaches to exploring the complexity of mycorrhizal responses.</jats:p>

Smoothing and extraction of traits in the growth analysis of noninvasive phenotypic data

“…Taking into account all of the plots presented in Additional file 1, log smoothing with six smoothing degrees of freedom is chosen for representing the PSA trend over DAPs 17-51. This differs from the previously reported analysis [11], where direct smoothing with six smoothing DF were used because the statistical analyses focused on DAP 27-43. To confirm this choice, a comparison of the results of direct and log smoothing for PSA, PSA AGR and PSA RGR when six smoothing DF are used is made using Additional file 1: Figures S4-S6, with the comparison for PSA AGR and PSA RGR also shown in Fig.…”

“…For example, Fig. 2a is presented, in a different form, without loess curves and without the outlier, in Figure 3 in Watts-Williams et al [11]; profile plots with loess curves are given in Figure 1 in Al-Tamimi et al [2].…”

“…responses per se or as a precursor to genetic analysis [1][2][3][4][5][6][7][8][9][10][11][12][13]. Because it involves non-invasive phenotyping of the same plants at different time points, it is now possible to measure many plants at many time points so that the precision of growth analyses [14] is much improved.…”

“…To illustrate the two techniques, the data from a tomato (Solanum lycopersicum) experiment [11] is used. This experiment involved the eight combinations of four levels of zinc (Zn) addition (0, 10, 40, and 90 mg Zn kg −1 soil) and either the addition of an arbuscular mycorrhizal fungal (AMF) inoculum (+AMF) or of a mock inoculum (−AMF) to the soil in pots with a single plant.…”

“…The experiment consisted of 32 potted plants that were placed in carts on the conveyor system within a Smarthouse (Australian Plant Phenomics Facility, University of Adelaide), where they were imaged daily from 17 to 51 days after planting (DAP). While the previously reported results [11] cover only 27-43 DAP, the full data set was processed and is the subject of the analyses reported here. 'Cart' will be used as the generic term for the unit in this experiment, each physical cart holding a pot with a single plant.…”