INTERNATIONAL JOURNAL OF AGRICULTURE & BIOLOGY ISSN Print: 1560–8530; ISSN Online: 1814–9596 11–109/MFA/2011/13–6–857–864 http://www.fspublishers.org Full Length Article To cite this paper: Apuan, D.A., M.A.G. Torres, M. Casimero, L.S. Sebastian and C.G. Demayo, 2011. Describing phenotypic variability in seed shapes of weedy rice types in comparison to cultivated and wild rice types using elliptic fourier analysis. Int. J. Agric. Biol., 13: 857–864 Describing Phenotypic Variability in Seed Shapes of Weedy Rice Types in Comparison to Cultivated and Wild Rice Types Using Elliptic Fourier Analysis DENNIS A. APUAN, MARK ANTHONY J. TORRES†, MADONNA CASIMERO‡, LEOCADIO S. SEBASTIAN‡ AND CESAR G. DEMAYO 1 † Department of Agricultural Sciences, Xavier University, Cagayan de Oro City, Philippines †Biological Sciences Department, College of Science and Mathematics, MSU-Iligan Institute of Technology, Iligan City, Philippines ‡Philippine Rice Research Institute, Science City of Munoz, Nueva Ecija, Philippines 1 Corresponding author’s e-mail: cgdemayo@gmail.com ABSTRACT Weedy rice is a serious threat to food security in a global scale. They invade lowland ricefields by having intermediate phenotypes between rice cultivars and its wild type. They grow sympatrically with the cultivars and compete effectively with the crop that often result to excessive yield loss, but un-fortunately this pest is difficult to control due to phenotypic resemblance and close genetic relationships with the rice cultivars. The high variability of weedy rice phenotype in the field is suggestive also that it reflect phenotypic relationships to its wild ancestors. In the current study, we explore the phenotypic affinity of weedy rice in the Philippine archipelago using seed shape. The shape is known to have large genetic bases and so its utility in the study is reliable. Using the Geometric Morphometric (GM) tool specifically elliptic Fourier analysis (EFA) and Multivariate Analysis in statistics, we found that 64% of the weedy rice in the archipelago has phenotypic affinity to 13 wild landraces (AA genome) collected from 15 different locations within West Africa, Carribean Islands, Latin America, India, Australia, South Asia and Southeast Asia. Ten populations have affinity to O. meyeriana (GG genome) in the Philippines and Malaysia. Both weedy populations from Misamis Oriental, Philippines (WRMIS1) and Nueva Ecija, Philippines (WRNE2) have affinity to PsBRc 64 and PSBRc 82, respectively while two populations from Iloilo, Philippines (WRILO1 & WRILO2) have affinity to O. latifolia in Costa Rica. Overall results display a complex pattern of phenotypic affinity, thus suggesting multiple origins of weedy rices in the Philippines. © 2011 Friends Science Publishers Key Words: Describing phenotypic variability; Cultivated; Wild rice; Elliptic fourier analysis INTRODUCTION Weedy rice has become a serious problem in the ricefield in a global scale (Ferrero et al., 1999; Mortimer et al., 2000; Cao et al., 2006; Delouche, 2007). In the Philippines, infestation rate is from 1% to 48% (Baltazar & Janiya, 2000). Yield reduction in rice reaches up to 80% (Smith, 1988). This pest is said to be associated to commercial rice varieties planted by direct seeding and broadcasting techniques (Sato, 2000; Delouche et al., 2007). Generally, its form is intermediate between cultivated and the wild types (Groot, 2003; Cao et al., 2006; Vaughan, 2008). In a study that attempted to draw the relationships of weedy rice with the wild types, only the Oryza rufipogon and O. nivara have so far are used most often as representative wild types to be compared to other wild rice types reported to be weedy in various geographical locations such as Oryza glaberrima, O. barthii, O. punctata and O. longistaminata in Africa (Delouche et al., 2007); as well as O. officinalis and O. meyeriana in China (Sato et al., 2000). Comparison has always been done qualitatively on the gross morphology of the rice plant such as the use of agronomic characters (Federici et al., 2002; Delouche et al., 2007), but not very successful. For example, in Costa Rica, Arrieta (2004) used agronomic characters to infer relationships of weedy rice with the commercial rice varieties and its wild type, but failed to detect small variations especially in shapes of the seeds. Structural shape of organisms has a large genetic basis (Yoshioka et al., 2004). It is a quantitative character involving many genes, thus variation in shape reflects underlying population genetic structure (Garnier et al., 2005). The utilization of a shape character in determining phenotypic affinity therefore apparently has large basis and high reliability. The shape of seeds of weedy rice are difficult to distinguish and essentially impossible to separate