Crop simulation models used for research or management, frequently require the development of leaf area to be mathematically described. This work examined, for maize, the relationship between the fully expanded area-per-leaf and leaf position, in an attempt to identify appropriate functions for use in such models. The slightly skewed bell-shaped function of the form LA(n) = Y0*exp(a* (xn - X0)2 + b* (xn-X0)3), was used to relate area-per-leaf (LA(n)) to leaf position (xn). The amplitude of the curve (Y0) represents the size of the largest leaf (at position X0). Plants varied in total leaf number (TLNO) from 12 to 17 and this variation had a major bearing on the total leaf area per plant (TLA). Both Y0 and X0 were linearly related to TLNO (r2 of 0.94 and 0.99 respectively). The parameter which controls the width of the bell-shaped curve (a), was non-linearly related to TLNO (r2 = 0.99), While the parameter that controls the degree of skewness (b), varied little with TLNO. Use of the bell-shaped curve in combination with the functions relating its parameters (Y0, a, X0) to TLNO, resulted in a root mean standard deviation (RMSD) of estimates of total leaf area per plant (TLA) of 176 cm2 for plants varying in TLNO from 12 to 17 and in TLA from 2839 to 6925 cm2. Bell-shaped functions were successfully fitted to independent area-per-leaf data from plants whose TLNO ranged from 13 to 26. The parameters developed in this study allowed area-per-leaf and TLA to be modelled accurately when non-limiting cultural conditions maximized leaf size. Under such circumstances, the bell-shaped curve with parameters linked to TLNO was shown to be a simple and flexible model of the potential area of individual leaves in maize. Resource constraints (such as those that can develop under high plant population) were shown to limit leaf size to less than that predicted by this model. Future work needs to examine the use of the bell-shaped curve in a simulation model context to address these constraints.
