International Society for Tropical Root Crops (ISTRC) 117 Assessing Potato Yellow Vein Virus (PYVV) infection using remotely sensed data and multifractal analysis P. Chávez 1 , P. Zorogastúa 1 , C. Chuquillanqui 2 , L.F. Salazar 2 , C.Yarlequé 1 , A. Posadas 1 , O. Piro 3 , J. Flexas 4 , V. Mares 1 , R. Quiroz 1 1 International Potato Center (CIP). Production Systems & the Environment Division. 2 International Potato Center (CIP). Integrated Crop Management Division. 3 University of Balearic Islands (UIB). Research Group on Plants under Mediterranean Conditions. Palma de Mallorca, Spain. 4 University of Balearic Islands (UIB). Physics Department. Palma de Mallorca, Spain. perla.chavez@uib.es , p.zorogastua@cgiar.org ; c.chuquillanqui@cgiar.org ; lsalazar@agdia.com ; c.yarleque@cgiar.org ; a.posadas@cgiar.org ; piro@ifisc.uib.es ; jaume.flexas@uib.es ; v.mares@cgiar.org ; r.quiroz@cgiar.org Abstract Potato Yellow Vein Virus (PYVV) reduces potato production in South America. Visual crop monitoring is a standard practice but the disease is generally detected after significant damage has occurred to photosynthetic tissues. Thus, a method for monitoring crop condition at different spatial scales to detect the disease before yields are severely affected is needed. Remotely sensed multispectral reflectance, based on the reflectivity and propagation of light inside plant tissues, was tested for the detection of PYVV infection in potato plants grown indoors. Visual assessment of symptoms in both virus-infected and healthy virus-free plants was compared to monitoring based on spectroradiometry and multispectral photographic images of the plants. Observed disturbances in light reflection by vascular tissues in infected plants, as well as spectral Vegetation Indices such as NDVI, SAVI, and IPVI, provided early detection of viral infection, long before symptoms of chlorosis were visually detected. To improve the earliness of detection, the raw remote sensing data from two subsequent experiments was further processed by multifractal analysis, a mathematical tool that addresses the scale dependency and variability of geophysical variables. Results showed that early diagnosis of PYVV infection was improved, providing the earliest detection of infection ever reported. For the first data set, the infection was evidenced 12 days after inoculation (23 days before the visual assessment did see the symptoms). For the second data set, our analysis denoted the disease 4 days after inoculation (33 days prior to the appearance of symptoms). Introduction Potato Yellow Vein Virus (PYVV), a Crinivirus of the Closteroviridae family (Salazar, 1998; Salazar et al., 2000), is considered a threat to potato production in South America (Saldarriaga et al., 1988; Salazar, 1998). PYVV is transmitted semi-persistently by 1) the whitefly Trialeurodes vaporariorum Westwood (Diptera, Sternorrhyncha, Aleyrodidae) (Díaz et al., 1990; Tamayo and Navarro, 1984; Salazar, 1998; Salazar et al., 2000), 2) tuber seed (Díaz et al., 1990; Salazar, 1998) and 3) by aboveground and underground stem-grafts (Salazar, 1998). The virus does not affect the size and morphology of plants; the typical yellowing of the veins appears between 30 and 40 days after infection, when stems and leaves are already formed (Saldarriaga et al., 1988). Research carried out on sugar beet (Beta vulgaris) also infected with a closteroviridae, the Beet Yellow Virus (BYV), showed damage in the photosynthetic mechanism of plants resulting in the reduction of net photosynthesis (Clover et al., 1999). One effect of the virus was the reduction of stomatal conductance, and a 50% reduction of the splitting of water in isolated chloroplasts (Spikes and Stout, 1955) with thylakoidal membrane damage (Baker and Horton, 1988; Salazar, 1998). Similar effects have been observed with PYVV in potatoes by Saldarriaga et al. (1988). Unpublished research at CIP shows that PYVV does infect the phloem cells and prevents the flow of carbohydrates, reducing yields drastically. PYVV infections are not evenly distributed across the field but they may occur in patches in current-season infections. A most widely used practice in the control of PYVV infection is the visual determination and manual elimination or roguing of infected plants and the spraying of pesticides over the whole stand at different times during the cultivation cycle, for the control of vectors. However, the use of pesticides increases costs and pesticide residue levels in agricultural products and soils, and contributes to ground water contamination.