Need We Model the Dynamics of Individual Specific Leaf Area with Respect to Carbon Partitioning?.

Leaf development of maize has been modeled by strong temperature response, but often neglected leaf carbon partitioning process (Tardieu et al., 1999). As the potential area of each leaf depends on its rank, the leaf area of individual leaf had to be calculated in separation, but the amount of carbon partitioned into the leaf, or the resultant biomass was not usually assessed on the leaf level (Fournier and Andrieu, 1998; Lizaso et al., 2003). Similarly, specific leaf area (SLA), which is an important indicator of leaf growth, was only taken account at the whole-plant level. In this research, we present that SLA of individual leaf calculated by the current maize growth model can be significantly different depending on how carbon assimilates are partitioned among the growing leaves.

A mechanistic maize simulation model, MAIZSIM, was used to calculate specific leaf area of individual leaves throughout the entire growth cycle (Kim et al., 2012). The leaf expansion was driven by thermal time accumulation and considered no limitation from water and nitrogen. Three carbon partitioning schemes were implemented to show leaf level SLA difference. The first was dependent on the constant potential leaf area, the second was dependent on the relative expansion rate of the leaf, and the last was equal distribution among the leaves.

The three partitioning schemes resulted into the exactly same leaf area throughout the growth period. When the grain filling started at Aug 4, 2007, the total leaf area was the exactly same 4278 cm² g^-1 and the total leaf biomass was 18.7, 18.9, and 18.9 g m^-2, respectively. The corresponding SLA for the total canopy was almost identical 228, 226, and 226 cm g^-1. These slight differences are probably due to complicated interactions between model components, but do not indicate a meaningful discrepancy due to different partitioning schemes. 

On the other hand, the SLA distribution pattern of individual leaf turned out to be very different between partitioning schemes. In the first case where the sink strength was determined by the potential leaf area which remained constant throughout the growth, individual SLA was stabilized at slightly above 200 cm g^-1. It looks quite reasonable only because the very same potential leaf area is used for the leaf area expansion, however, we do not have mechanistic background that the leaf carbon partitioning follows the potential leaf area which is independent of carbon assimilation process. The second case is where the source strength is dependent on the relative expansion rate of the leaf. Though the magnitude of individual SLA values seems to be too high for reality, it was the only case that at least captured a descending gradient of SLA from the top to the bottom of the canopy. It is known that SLA distribution often inversely correlates with irradiance extinction pattern where photosynthetic capacity and thus corresponding nitrogen profile is maximized at the top. The last case is where the source strength is only dependent on the number of leaves so that always an equal portion of carbon is allocated regardless of the size or age. It looked like somewhere between the first and the second cases where individual SLA was not too high as the second case, but might still need some compensation. Its overall bell-shaped profile of SLA resembled the first case.

We showed that different carbon partitioning schemes could drive radically different growth dynamics at the individual leaf level, but still contribute the exactly same at the whole plant level. We do not suggest any particular carbon partitioning scheme here, however, rather calling for an attention that the current practice of modelling leaf development independent of carbon partitioning would miss another important dynamics of the crop growth.