ABSTRACT

Maize (Zea mays L.) is adapted to all agro-ecologies in west Africa. Varieties of different maturity groups are needed to maximally exploit the agro-ecologies for maize production. Silking date is used to quantify maturity but this is affected by G x E interaction (GEI) and, therefore, unreliable. We hypothesized that, in response to climatic factors, maize varieties have genetic constants that are unaffected by GEI. The objectives of this study were to (i) develop and validate models containing temperature, heat units and photoperiod for predicting flowering time and (ii) classify maize varieties into maturity groups using the best model. One hundred maize varieties were evaluated at Ile-Ife, Nigeria for 2 yr under two contrasting seasons each yr. Days to flowering (f), temperature and photoperiod were used to develop two models for individual varieties: 1/f = a + bX for X = temperature, photoperiod, or heat units; and 1/f = a + b₁X₁ + b₂X₂ for temperature and photoperiod. Models developed from the 2007 late season data were validated by computing the predicted number of days to flowering for the other seasons, using long term mean values for the climatic variables and comparing with the observed. Days to anthesis (DA with mean = 60.6±0.43) was the best flowering trait for quantifying maturity and the thermal (temperature) model (mean = 58.6±0.92) best predicted the trait. For the observed DA, the varieties were classified into maturity groups as follows: late 50 varieties, intermediate 18, early 25 and extra-early 7. Corresponding classification based on the predicted values from the thermal model was 24, 27, 30 and 19 varieties, respectively. Number of varieties common to both classifications was 23, 4, 15 and 7 respectively, a total of 49%. With further refinements, the thermal model may be reliably used to classify maize into maturity groups.