Keynote Speakers and tentative title

Christine Beveridge: Using prior knowledge to build predictive models for plant success

ARC Centre for Plant Success in Nature and Agriculture, University of Queensland

Tentative abstract: The plant phenotype is the emergent property of a huge number of interactions among environment, life history, internal signals (e.g., hormones) and genome. For this reason it is challenging to predict the phenotype of plants. We seek a step-wise manner in which to use diverse prior knowledge to build models that can enhance our ability to predict plant phenotype for benefit in fundamental science, crop improvement and the understanding of evolution. This presentation will focus on our related work in modelling shoot branching.

Short Bio: Prof Christine Beveridge graduated with a BSc (Hons) and PhD in Plant Science at the University of Tasmania in 1994.  After a two-year Postdoctoral Fellowship at the National Institute of Agricultural Research (Versailles, France) Christine took up a competitive University of Queensland (UQ) Fellowship, then an Australian Research Council (ARC) Postdoctoral Fellowship, teaching and research positions, an ARC Future Fellowship and Deputy Dean and Associate Dean Research (Science) position at UQ.  Christine is now the Director of the ARC Centre of Excellence for Plant Success in Nature and Agriculture (CoE), a Fellow of the Australian Academy of Science, an ARC Georgina Sweet Laureate Fellow and a highly cited researcher.  Christine was both the first female and first Australasian president of the International Plant Growth Substances Association and is a Life Member of the Australian Society of Plant Scientists.  Christine discovered strigolactone as a plant hormone and that sugar signalling is a driver of shoot branching.  Christine founded the CoE of 175 members including 17 Chief Investigators across five Universities to enhance the use of prior knowledge in breeding, enhance food diversity, and to promote an inclusive interdisciplinary research culture to future proof Australian agriculture, biodiversity and sustainability.

Twitter account: https://twitter.com/cabeveridge29

List of three indicative publication:

Powell, O. M., F. Barbier, K. P. Voss-Fels, C. Beveridge and M. Cooper (2022). Investigations into the emergent properties of gene-to-phenotype networks across cycles of selection: A case study of shoot branching in plants. in silico Plants 4(1): diac006.

Bertheloot J, Barbier F, Boudon F, Perez‐Garcis MD, Péron T, Citerne S, Dun E, Beveridge C, Godin C, and Sakr S (2020). Sugar availability suppresses the auxin‐induced strigolactone pathway to promote bud outgrowth. New Phytologist 225:866-879.

Dun EA, Hanan J, Beveridge CA (2009) Computational modeling and molecular physiology experiments reveal new insights into shoot branching in pea. Plant Cell 21: 3459-3472.

Andrew French: Data is everything: the power of plant images in the age of deep learning

Computer Vision Lab, University of Nottingham

Tentative abstract: Deep learning has unquestionably changed the way we use predictive modelling in plant science. For ten years, deep models in the form of convolutional neural networks have been conceived, developed and refined, to the point where we now have a suite of powerful models for most common phenotyping tasks. But the problem is not solved. The more powerful the model, the greater the demands on the datasets on which they must be trained. This has never been more true than with the latest advances in transformer models, which are largely pretrained on enormous datasets. This talk will concentrate on the data side of the deep learning coin, and will look at some novel ways of boosting data set sizes we have developed at Nottingham.

Short Bio:  Andrew is a Professor of Computer Science at the University of Nottingham, where he co-leads the Computer Vision Laboratory.   For 15 years he has worked closely with the plant science community, developing software tools and new data analysis approaches, looking at plants from the cell to the field scale.  Deep machine learning drives much of the image analysis in the lab today.  Andrew currently leads a project developing data science training for plant phenotyping (DataCAMPP – Training in Data Capture, Analysis and Management for Plant Phenotyping).

List of three indicative publication:

Ganana: unsupervised domain adaptation for volumetric regression of fruit. ZKJ Hartley, AS Jackson, M Pound, AP French. Plant Phenomics (2021)

Domain adaptation of synthetic images for wheat head detection. ZKJ Hartley, AP French. Plants 10 (12), 2633 (2021)

Learning to Localise and Count with Incomplete Dot-annotations. F Chen, MP Pound, AP French. Proceedings of the IEEE/CVF International Conference on Computer Vision (2021)

Anja Geitmann: Plants in motion - using engineering methods to understand plant morphogenesis and actuation

McGill University

Tentative abstract: During the plant's life cycle, organs and tissues change shape and deform both elastically and plastically. Permanent shape change is characteristic of plant growth processes whereas reversible (elastic) behavior is typically associated with responses to environmental triggers. Both behaviors rely on mechanical processes involving the cell wall and turgor pressure. Through computational modeling and micromechanical testing we interrogate the mechanical underpinnings of these processes.

Short Bio: Dr. Geitmann’s research focuses on the cellular processes driving plant reproduction and morphogenesis. She holds the Canada Research Chair in Biomechanics of Plant Development and leads an interdisciplinary team of biologists and engineers. Her research combines cell biology with high-end imaging, micromanipulation, and computational modeling to study the mechano-structural underpinnings of plant functioning. Dr. Geitmann has served as the President of the International Association of Plant Reproduction Research, the Microscopical Society of Canada and the Canadian Society of Plant Biologists. She serves on the editorial boards of multiple scientific journals including Cell and Plant Physiology.

Twitter account: https://twitter.com/GeitmannLab 

List of three indicative publication:

Sleboda DA, Geitmann A. Sharif-Naeini R. 2022. Multiscale structural control of hydraulic bending in the sensitive plant Mimosa pudica. bioRxiv doi.org/10.1101/2022.02.28.482281

Bidhendi AJ, Altartouri B, Gosselin FP, Geitmann A. 2019. Mechanical stress initiates and sustains the morphogenesis of wavy leaf epidermal cells. Cell Reports 28: 1237-1250

Bidhendi AJ, Geitmann A. 2018. Finite element modeling of shape changes in plant cells. Plant Physiology 176: 41-56

Tracy Lawson: Phenotyping photosynthesis and stomatal kinetics in leaf and non foliar tissue

University of Essex

Tentative abstract: In order for leaf photosynthesis to take place CO2 must enter the leaf through adjustable pores, called stomata, and at the same time water is lost through these pores which also aids in cooling of the leaf. As stomatal behaviour controls photosynthesis, water loss and leaf temperature these pores are an unexploited but important targets for manipulation to improve crop productivity. Stomata are found on both leaf and non-leaf material in significant and differential numbers. Here we use chlorophyll fluorescence imaging and thermograph to explore the role of stomata in gaseous exchange and evaporative cooling for improved photosynthesis under different climatic conditions.

Short Bio: Tracy is Group Convener, Director of Plant Phenotyping and Director of Impact at Essex, with over 25 years’ experience in photosynthesis research. Her research focuses on the stomatal control of atmospheric gas entry into the leaf, associated water loss and the mechanisms that regulate this process. Recent research has paid particular attention to stomatal kinetics and the impact of dynamic environments on both photosynthesis and stomatal behaviour. Tracy’s work also concentrates on phenotyping including chlorophyll fluorescence techniques (for quantifying light use and photosynthetic efficiency) and thermal imaging (for measuring stomatal responses and kinetics). Lawson’s lab developed the first imaging system for screening plant water-use-efficiency (McAusland et al., 2013).

Twitter account: https://twitter.com/drtracylawson 

Lawson, T. Vialet-Chabrand S. (2019). Speedy stomata, photosynthesis and plant water use efficiency. Invited Tansley Insight. New Phytologist 221: 93-98

Simkin AJ, Faralli M, Ramamoorthy S, Lawson T. (2020). Photosynthesis in non-foliar tissues: implications for yield. The Plant Journal 101: 1001-1015

Wall S, Vialet-Chabrand S., Davey P, van Rie J, Galle A, Cockram J and Lawson T. (2022) Stomata on the abaxial and adaxial leaf surface contribute differently to leaf gas exchange and photosynthesis in wheat. New Phytologist.235:1743–1756

Leo Marcelis: Crop models in controlled environment agriculture: exploring the variation within canopies

Wageningen University, The Netherlands

Tentative abstract: This presentation will describe some approaches for mechanistically simulating growth, development and plant architecture of greenhouse-grown crops in response to abiotic constraints. Some examples of exploring the variability of light distribution and photosynthesis within canopies will be shown; this will include examples of different row orientation or different types of LED lighting. It will be discussed how these models may be used for prediction and planning of production, decision support systems, control of the greenhouse climate, supply of water and nutrients, and phenotyping. It will be discussed how models help us in exploring new avenues, and how the combination of models and sensors is powerful in both monitoring and phenotyping.

Short Bio: Prof dr Leo Marcelis is head of the chair group Horticulture and Product Physiology at Wageningen University. Leo has a a strong background in plant physiology, crop monitoring, computational modelling and experimentation. He has extensively studied the physiology, growth and development of plants in order to improve sustainability and quality of crop production in greenhouses and vertical farms. In particular fluxes of assimilates, water and nutrients in the plant, sink/source interactions and partitioning among plant organs in response to abiotic conditions (including LED lighting) are subject of study.

Twitter account: https://twitter.com/leomarcelis 

List of three indicative publication:

Schipper, R., van der Meer, M., de Visser, P.H.B., Heuvelink, E.,and Marcelis L.F.M. 2023. Consequences of intra-canopy and top LED lighting for uniformity of light distribution in a tomato crop. Frontiers in Plant Science 14:1012529

Van Der Meer, M., De Visser, P. H. B., Heuvelink, E. & Marcelis, L. F. M. 2021. Row orientation affects the uniformity of light absorption, but hardly affects crop photosynthesis in hedgerow tomato crops. in silico Plants. 3, 2, diab025.

Zhang, N., Van Westreenen, A., Evers, J. B., Anten, N. P. R. & Marcelis, L. F. M. 2020. Quantifying the contribution of bent shoots to plant photosynthesis and biomass production of flower shoots in rose (Rosa hybrida) using a functional–structural plant model. Annals of Botany. 126, 4, p. 587-599

Christopher McCool: Exploiting spatial-temporal information in robotic vision systems: with applications to agriculture

University of Bonn

Tentative abstract: Robotic platforms are designed to traverse an environment in order to automate tasks such as crop or plant monitoring. Current approaches to crop monitoring are dominated by deep learning approaches that generally use snapshots, or still images, of the scenes. Despite the success of these approaches there are also major limitations. In this presentation, I will discuss recent developments in the robotic vision systems at the University of Bonn which exploit the fact that we have a robot moving in the field. This enables us to exploit a rich set of both spatial and temporal information, available from robots, which can be used to enhance vision systems to provide greater and more accurate information.

Short Bio: Chris is a Professor at the University of Bonn and leads a group which develops robotic vision algorithms that enables robots and autonomous systems to work in challenging environments (especially agriculture). Prof. McCool received his PhD in 2007 from the Queensland University of Technology (QUT), Australia. He has previously worked at the Idiap Research Institute (Switzerland), National ICT Australia and the Queensland University of Technology (Australia).

List of three indicative publication:

C. Smitt, M. Halstead, A. Ahmadi, and C. McCool, “Explicitly incorporating spatial information to recurrent networks for agriculture}”, accepted to Robotics and Automation Letters, 2022.

M. Halstead, A. Ahmadi, C. Smitt, O. Schmittmann, and C.McCool, “Crop Agnostic Monitoring Driven by Deep Learning”, Frontiers in Plant Science, 2021.

A. Ahmadi, M. Halstead, and C. McCool, “Virtual Temporal Samples for Recurrent Neural Networks: applied to semantic segmentation in agriculture”, GCPR/DAGM, 2021.

Christophe Pradal: Data-intensive scientific workflows for model-assisted high-throughput phenotyping

CIRAD & inria, Montpellier

Tentative abstract: High-throughput phenotyping platforms allow the study of the form and function of a large number of genotypes subjected to different growing conditions (GxE). Automatic computational pipelines for phenotyping are able to characterize the structure and the development of plants at an unprecedent resolution. Scientific workflows are way to schedule these complex pipelines on distributed cloud infrastructure, to manage the huge amount of data and to enhance the reproducibility of such experiments. In this presentation, I will discuss the recent developments in root and shoot phenotyping methods, how it challenges FSPM formalisms and platforms, and how scientific workflows management system can help to improve the connection between phenotyping and modelling communities while reducing the processing and environmental cost of the computation.

Short Bio: Christophe is a Senior Researcher at CIRAD, Montpellier and an associate researcher at inria. He co-leads the interdisciplinary group PhenoMen at the crossroads of Data science (modeling & phenotyping), ecophysiology and agro-ecology in the AGAP Institute with Christine Granier. In the last 20 years, he has worked in the FSPM community, leading the OpenAlea platform, and designing models, algorithms and data structures in plant phenotyping and modelling at different scales.  He served during 5 years as an associate editor of Plant Methods.

Twitter account: https://twitter.com/agapinstitut #PhenomenTeam

List of three indicative publication:

G. Heidsieck, D. De Oliveira, E. Pacitti, C. Pradal, F. Tardieu, P. Valduriez (2021). Cache-aware scheduling of scientific workflows in a multisite cloud. Future Generation Computer Systems, 122, 172-186.

B. Daviet, R. Fernandez, L. Cabrera-Bosquet, C. Pradal*, C. Fournier* (2022). PhenoTrack3D: an automatic high-throughput phenotyping pipeline to track maize organs over time. bioRxiv.

H. Takahashi, C. Pradal (2021). Root phenotyping: important and minimum information required for root modeling in crop plants. Breeding Science, 71(1), 109-116.

Przemyslaw Prusinkiewicz: Experimental data, computational models, and theorems about plants

University of Calgary

Tentative abstract: I will present a perspective on the role of models and mathematical reasoning in studies of plant development, focusing on the multiple connections between biology and geometry. The emphasis will be on the 20th and 21st century mathematics that significantly extends concepts already recognized as relevant to biology, such as fractals and L-systems. In this context, I will show how the theories of quasicrystals, aperiodic patterns and point distribution (the three gap theorem) are contributing to our understanding of the phyllotaxis and structure of flower heads. Interestingly, some precursors of these theories appeared in biological literature before they became the subject of mathematical studies that are now used to further advance biology.

Short Bio: Przemyslaw Prusinkiewicz is a Professor Emeritus of Computer Science at the University of Calgary, Canada. He is a pioneer of computational modeling, simulation and visualization of plant development. His current research is focused on computational models of development that link molecular-level processes to macroscopic plant forms. Professor Prusinkiewicz is a Fellow of the Royal Society of Canada, a honorary member of Polish Botanical Society, and a recipient of the Association for Computing Machinery SIGGRAPH Achievement Award and the Canadian Human Computer Communications Society Achievement Award for his work on plant modeling.

Twitter account: https://twitter.com/pprusinkiewicz 

List of three indicative publication:

T. Zhang, M. Cieslak, A. Owens, F. Wang, S. K. Broholm, T. H. Teeri, P. Elomaa, P. Prusinkiewicz. Phyllotactic patterning of gerbera flower heads. Proceedings of the National Academy of Sciences USA 118(13), e2016304118, 2021.

P. Prusinkiewicz, T. Zhang, A. Owens, M. Cieslak, P. Elomaa. Phyllotaxis without symmetry: what can we learn from flower heads? Journal of Experimental Botany, erac101, 2022.

J. Battjes, P. Prusinkiewicz. Modeling meristic characters of Asteracean flowerheads. In R. V. Jean and D. Barabé (Eds.): Symmetry in Plants, World Scientific, Singapore, 1998, pp. 281-312.