Effect of temperature and photoperiod on the phenological development of wild mustard (Sinapis arvensis L.) Jian Zhong Huang a , Anil Shrestha a , Matthijs Tollenaar a , William Deen a , Irena Rajcan a , Hamid Rahimian b , Clarence J. Swanton a,* a Department of Plant Agriculture, University of Guelph, Guelph, Ont., Canada N1G 2W1 b Department of Agronomy, College of Agriculture, Mashhad, Iran Received 20 May 2000; received in revised form 10 December 2000; accepted 14 December 2000 Abstract The development of mechanistic weed models focuses on determining the outcome of weed±crop interference. Phenological development is a major factor determining the outcome of weed and crop competition. The in¯uence of temperature and photoperiod on phenological development of wild mustard (Sinapis arvensis L.) was studied in growth cabinets. The life cycle of wild mustard was de®ned in terms of biological days (Bd: chronological days at the optimum photoperiod and temperature). Wild mustard was a long-day species adapted to a wide temperature range of 1.5±488C. Four phases of development of wild mustard were described: (1) a juvenile phase of 12.7 Bd; (2) a photoperiod-sensitive inductive phase of 6.2 Bd; (3) a photoperiod-sensitive post-inductive phase of 12.8 Bd; (4) a photoperiod-insensitive phase of 33.9 Bd. When effects of photoperiod on rate of development were normalized across phases of development, photoperiod sensitivity did not vary among phases of development. Interpretation of constant sensitivity to photoperiod will simplify simulation of weed phenology in mechanistic models. # 2001 Published by Elsevier Science B.V. Keywords: Phenology; Mechanistic weed models; Juvenile phase; Biological days 1. Introduction The use of predictive approaches is an important component of an integrated weed management (IWM) system (Swanton and Murphy, 1996). An understand- ing of weed phenology is important for such an approach and for the development of predictive mechanistic models (Elmore, 1996). Two types of models have been developed for the prediction of weed occurrence: mechanistic and empirical models. Mechanistic models are usually better than empirical models in assessing the impact of genetic variation and environmental factors on the development pro- cesses of the crop and weed (Ghersa and Holt, 1995). Empirical models cannot account for these variations and are therefore of limited predictive ability (Kropff et al., 1992; Swanton and Murphy, 1996). Mechanistic models integrate our current understanding of pheno- logical development of both crop and weed under competition. Phenological development of weeds such as green foxtail (Setaria viridis L.) (Swanton et al., 1999), barnyardgrass (Echinochloa crus-galli (L.) Beauv.) (Swanton et al., 2000), and common lambs- quarters (Chenopodium album L.) (Huang et al., 2001) have been described previously. Field Crops Research 70 (2001) 75±86 * Corresponding author. E-mail address: cswanton@plant.uoguelph.ca (C.J. Swanton). 0378-4290/01/$ ± see front matter # 2001 Published by Elsevier Science B.V. PII:S0378-4290(00)00155-6