High-Throughput Automated Phenotyping As A Shortcut To More Effective Biostimulants: From Seeds to Crops

Mirella Sorrentino1,4, Nuria de Diego2, Giuseppe Colla3, Lukáš Spíchal2, Youssef Rouphael4 and Klára Panzarová1

1PSI (Photon Systems Instruments), spol. s r.o., Drasov, Czech Republic

2Department of Chemical Biology and Genetics, Centre of the Region Haná for Biotechnological and Agricultural Research, Faculty of Science, Palacký University, Olomouc, Czechia

3 Department of Agriculture and Forest Sciences, University of Tuscia, via San Camillo De Lellis snc, 01100 Viterbo, Italy

 4Department of Agriculture, Università of Naples Federico II, via Università 100, 80055 Portici (NA), Italy


Development of highly effective biostimulants requires an accurate evaluation of the effects of candidate products on morpho-physiological traits of selected crops during different developmental stages and environmental conditions. As conventional screening methods are time consuming, destructive and labour intensive, high-throughput plant phenotyping procedures were recently proposed as effective and high-precision tools for novel product screening. Plant phenotyping systems are fully automated robotic systems usually installed in a controlled environment or in semi-controlled greenhouse conditions. The phenotyping platforms are designed to ensure not only non- invasive image-based monitoring of plants in throughput of few up to several hundreds of plants, but also provide means for automated cultivation and handling of the plants such as automated watering/weighing, nutrient delivery or spraying units (1, 2). Plants can be dynamically scored for many morpho-physiological and biochemical traits related to growth, yield and performance throughout their development or onset, progression and recovery from abiotic stress. This approach thus allows for biostimulant functional characterization in planta in high-precision and high-resolution in a given phase of plant development and/or plant response to environmental conditions.

So-called integrative phenotyping approach in multi-sensoric phenotyping platforms has been recently used to succesfully investigate the mode of action of the biostimulants and individuate the morpho-physiological traits related to stress and biostimulant´s applications (2,3,4). The broad spectrum of traits can be quantitatively described and qualitatively differentiated by including imaging sensors not only for visible imaging (RGB imaging) and/or 3D imaging, but also for imaging spectroscopy (hyperspectral imaging), thermal infrared imaging and chlorophyll fluorescence imaging. RGB imaging is used to estimate the true colour of each pixel and by using image processing algorithms for identification of plant-derived pixels. For identified plant objects, morphological and geometrical features are quantified, including colour properties. The pixel number-based assessment of plant volume or total leaf area correlates with fresh and dry weight of above ground plant biomass. Plant growth can be captured in time-series measurements that are necessary to follow the progression of growth dynamics on individual plants in given phase of development and/or stress response. Precise architecture and shape of the plants such as height, individual leaf size or biomass can be quantified by using 3D imaging technology. In addition to structural and morphological characterisation key feature of the automated phenotyping platforms are sensors designed for so called physiological phenotyping. Kinetic chlorophyll fluorescence (ChlF) imaging is used to quantify plant´s photosynthetic capacity and ability to harvest light energy. ChlF is a popular technique in plant physiology used for rapid non-invasive measurement of photosystem II (PSII) activity. PSII activity is very sensitive to a range of biotic and abiotic factors and therefore the chlorophyll fluorescence technique is used as a rapid indicator of photosynthetic performance of plants in different developmental stages and/or in response to changing environment. The advantage of chlorophyll fluorescence measurements over other methods for monitoring stresses is that changes in chlorophyll fluorescence kinetic parameters often occur before other effects of stress are apparent. In addition thermal imaging is very sensitive technique used for measurement of stomatal conductance and water transpiration rates of plants by leaf and canopy temperature quantification. Thermal imaging of leaves is important in assessing plant’s responses to heat load and water deprivation. Finally hyperspectral imaging technology can be used for quantification of spectral reflectance profiles of the plants and characterisation of plant´s biochemical properties by non-destructive image-based technology. Hyperspectral images can be used to calculate range of reflective indices across the entire surface of the plant referring to pigment composition of the plant, nitrogen status or water content of the plants.

Here we present one example how the integrative phenotyping approach can be applied for screening of newly-formulated and commercial plant-derived protein hydrolysates (PHs). The set of PHs was screened for their putative growth promoting and/or stress alleviating function on plants subjected to an abiotic stress (salinity) at different phenological stages. Using two automated PlantScreenTM phenotyping platforms developed by PSI (Photon Systems Instruments, Czechia), we were able to monitor the mode of action of the selected PHs by using different mode of application in different plants species and screened in different phenological stages. PHs were applied by 1) seed coating, 2) in growing media or 3) by foliar spray. The effects of PHs was assessed by quantifying 1) seedling emergence rate of wheat, 2) rosette growth-related traits in Arabidopsis and 3) set of morpho-physiological traits in lettuce and tomato plants (Fig.1). The PHs biostimulant mode of action was characterised by quantitative analysis of photosynthetic performance, growth dynamics and colour analysis.

Fig.1| High-throughput screening pipeline for PHs functional characterisation. Biostimulant mode of action is quantified in different plant species throughout set of developmental stages and by using different mode of biostimulant application. The PHs activity was assesed from the seed phase using germination assay, going to seedling phase using in-vitro assay in Arabidopsis, up to leafy crops such as lettuce and fruiting crops such as tomato. The PHs were applied by seed coating (Germination assay), were mixed in media (In-vitro assay) both performed in OloPhen XYZ Platform (CR Hana Olomouc , Czech Republic) and by spraying of plants as performed in the PlantScreenTM Compact System (PSI Research Center, Drasov, Czech Republic).

The emergence assay and the in-vitro assay, were performed in the OloPhen XYZ Platform (Centre of the Region Haná for Biotechnological and Agricultural Research, Olomouc). The two assays were used for initial screening of larger set of substances in range of application concentrations and in different conditions. Using the two initial assays we aimed at selection of most promising candidate PH´s and best application doses for second phase in depth morpho-physiological mode of action characterisation of the selected substances in the crop plants. Briefly, in the emergence assay wheat seeds coated with 6 different biostimulants at increasing concentration were sown in soil saturated either with tap or salt-enriched water and their emergence rate and their growth were monitored for 12 days using an RGB camera. For the second trial, the in-vitro assay, 4-day-old Arabidopsis thaliana seedlings were grown for 8 days in 48-well plates containing a media enriched with 13 different biostimulants at increasing concentrations, either in control or salt stress conditions. An RGB picture of each plate was taken twice a day for the entire duration of the experiment. Green area increase, relative growth rate and the survival rate in salt conditions were used to calculate the Plant Biostimulant Characterization Index (PBC) (3) in order to identify the substances behaving as growth improvers and/or stress alleviators. Best performing substances were selected for the second phase screening in leafy crops performed in the PSI Research Center. Here we present first results for one of the substances characterised in initial assay as a stress alleviating compound (Fig.2). The lettuce plants were grown for 35 days in controlled conditions and were watered every second day either with a nutrient solution or with an 80 mM NaCl solution. The biostimulant was applied twice a week as foliar spray. Growth of the plants was quantified dynamically throughout development and salt imposition (Fig.2A). Growth performance was improved in salt grown plants sprayed with the biostimulant (Fig. 2B,C), which correlated with the fresh weight of the plants quantified at the end of the trial (Fig.2D) finally confirming that the selected substance acts as a stress alleviator.

Fig.2| Non-invasive image-based phenotypical analysis of control and PH´s treated lettuce plants grown in control or salt-stress conditions by using the PlantScreenTM Compact System. (A) Colour-segmented top view Red Green Blue (RGB) images of the lettuce plants over the time of phenotyping period, from the 10th to the 31st Day After Stratification (DAS). (B) Projected shoot area of lettuce plants cultivated in control or salt-stress conditions and subjected to treatment by biostimulant. Values are expressed as mm2 and represent the average of ten biological replicates per treatment ± standard deviation. (C) Projected shoot area of lettuce plants in the last day of phenotyping. Values represent the average of ten biological replicates per treatment. Error bars represent standard deviation. (D) Fresh weight of lettuce shoots harvested following the end of the phenotyping period (31 DAS). Values represent the average of ten biological replicates per treatment. Error bars represent standard deviation.

Given the promising results obtained, we will next investigate the mode of action of a broader set of biostimulants selected from the initial in vitro screen and will characterise their activity in leafy and fruiting crops, lettuce and tomato plants respectively.

In summary the pre-liminary data presented here show that the high-throughput automated phenotyping proved to be a powerful tool, allowing us to follow the plant performance throughout its life, in all developmental stages, individuate the morpho-physiological traits related to stress and biostimulants applications and investigate the mode of action of the substances. Precise and accurate assessment of phenotypic variables is critical to characterise and quantify the biostimulant activity of various products selected as putative growth improvers and/or stress alleviators. High-throughput phenotyping technologies that are slowly receiving increasing attention for the purpose of product screening and development might in future provide very time and cost efficient means to screen for novel substances, optimal doses, right time of application and mode of action of the compounds with biostimulants activities.