Fast and non-invasive phenotyping of plant health/stress status using a LED induced chlorophyll fluorescence transient imager Henk Jalink, Wageningen UR Greenhouse Horticulture, The Netherlands EPSO: The European Plant Science Organisation EPSO Workshop on Plant Phenotyping November 02-03, 2009 Forschungszentrum Jülich, Germany Forschungszentrum Jülich, Germany ICG-3: Phytosphere Jülich Plant Phenotyping Centre (JPPC) Website: http://www.jppc.de http://www.plantphenomics.com/phenotyping2009
Fast and non-invasive phenotyping of plant health/stress status using a LED induced chlorophyll fluorescence transient imager Henk Jalink Wageningen UR Greenhouse Horticulture, The Netherlands
What is needed: Technologies to characterize plant performance and dynamics High throughput techniques to measure plant at desired scenarios
What is needed: Technologies to characterize plant performance and dynamics High throughput techniques to measure plant at desired scenarios Objective: Develop an imaging sensor that monitors the health and dynamic response of whole plants
Advantage of chlorophyll fluorescence: only plant tissue is visible; no background LED imager
Spectral properties of chlorophyll Fluorescence a b Laser a fluorescence Camera Absorbance b Chlorophyll-a 400 450 500 550 600 650 700 750 800 Wavelength [nm]
Plant recognition using color Easy for computer to recognize the plant Difficult for computer to recognize the plant
Color and fluorescence image of a plant Difficult for computer to recognize the plant Easy for computer to recognize the plant on fluorescence
Color and fluorescence image of a plant Integration time of 0.1 s
MIPS TM robot arm with colour and CF camera Robot arm for scanning plants MIPS TM camera with laser for CF image Light source for colour image Table with plants
Outgrowth of Phytophthora on potato leaf in petri dish Basic fluorescence Maximal fluorescence Water control Photosynthetic activity Pathogen Calculated infected area (gray color)
Outgrowth of Phytophthora on potato leaf Research on pathogenisis Q.E. (%) 80 70 60 50 40 30 Start of growth Research on resistance of cultivars 50 40 30 20 10 Infected area (%) 20 0 0 15 30 45 60 75 90 105 120 135 150 Time (hrs) Growth rate
Herbicide formulation testing on PA 2 droplets of herbicide
Counting pixels with PA < 40% Number of pixels with PA< 40% 1400 1200 1000 800 600 400 200 With additive better uptake of herbicide Control Glyphosate Glyphosate+POE (15) tallowamine 0 0 20 40 60 80 Time after treatment (h)
Some conclusions MIPS system is too slow for high throughput and screening: about 20 sec MIPS can not measure at elevated light levels Needed: a fast methodology that is able to saturate the photosynthesis within a sec
LED Induced Fluorescence Transient Imager for Measuring Stress CCD-camera 14 bit LED s 5000 Watt 1000 µmol/m 2 s at 40x40 cm 2 Background light
Methodology of the LED Induced Fluorescence Transient Imager CF induction (Kautsky) curve of a corn leaf F(t) = F 0 + (F max -F 0 )*(1 e -t/τ ) Asymptote 50 successive images Linear on log-scale: approximate by exponential curve Sinsawat, 1999 PhD thesis ETH
Led pulse of 15 ms; good contrast 1
Led pulse of 15 ms 2
Led pulse of 15 ms 3
Led pulse of 15 ms 4
Led pulse of 15 ms 5
Led pulse of 15 ms 6
Led pulse of 15 ms 7
Led pulse of 15 ms 8
Led pulse of 15 ms 9
Hardly any contrast due to saturation of photosynthesis 10
Time response of healthy tissue Normal functioning photosynthesis
Time response of partly inhibited photosynthesis Partly inhibited photosynthesis
Time response of fully inhibited photosynthesis Fully inhibited photosynthesis
Creating images of PA and τ by fitting each pixel F(t) = F 0 + (F max -F 0 )*(1 e -t/τ ) F max Photosynthetic activity = (F max F 0 )/F max F 0 τ: time constant Two independent parameters from which images can be calculated
Image of the photosynthetic activity, PA High PA Healthy Low PA
Image of the time constant, τ Slow τ Healthy Fast τ
Robotized LED imager for phenotyping at different scenarios LED imager on robot arm Climate room Fully programmable 30 images within one second Calculation of PA- and τ- image Growth of projected leaf area Can run for weeks
Drought stress: measurement in the dark (adaptation) Control Drought Saintpaulia (African violet) PA Dark
Drought stress: measurement in the light (adaptation) Control Drought Saintpaulia (African violet) PA 90 µmol/m 2 s
Black nightshade stressed and healthy
Black nightshade healthy and heavily stressed PA τ Healthy PA τ Stressed
Wilting of a detached leaf of black nightshade plant PA PA PA PA t=0 t=15 min. t=30 min. t=60 min. τ τ τ τ At t=1 min. a leaf is detached
Screening potato genotypes on salinity stress τ Stressed Control
Summary Stress of whole plants can be visualized on PA and τ Changes in PA- and τ-image can depend on the type of stress Only stress that influences the photosynthetic apparatus like diseases, herbicides, light stress and drought Non-invasive, quantitative, accurate and objective Plants can be monitored continuously at different scenarios Within a time frame of 300 ms: high throughput
Future development New camera system for fluorescence and reflection imaging Camera with 8 position filter wheel White LED s for reflection imaging Pulsed high intensity red LED s for imaging Kautsky curve LED s for fluorescence imaging of chlorophyll and anthocyanins content
Time for discussion Dr. Henk Jalink Wageningen UR Greenhouse Horticulture Visiting address: Building 107, Droevendaalsesteeg 1, 6708 PB Wageningen Mail address: P.O. Box 644, NL 6700 AP Wageningen, The Netherlands Phone: +31 (0) 317 48 08 44 Fax: +31 (0) 317 41 80 94 Mailto: henk.jalink@wur.nl http://www.greenhousehorticulture.wur.nl