Submitted to: Ecological Modelling
Publication Type: Peer reviewed journal
Publication Acceptance Date: 5/12/2003
Publication Date: 10/1/2003
Citation: Ferreyra, A.R., Dardanelli, J.L., Pachepsky, L., Collino, D., Faustinelli, P., Giambastiani, G., Reddy, V., Jones, J.W. 2003. Nonlinear effects of water stress on peanut photosynthesis at crop and leaf scales. Ecological Modelling. 168(1-2):57-76. Interpretive Summary: Managing water regimes of peanut crop is an important part of farm practices in many regions. Crop model PNUTGRO has modules intended to simulate crop responses. Testing the PNUTGRO version 01 with data on growth and development of peanut cv. Florian INTA in three different locations in Argentina has shown that the model performs adequately for irrigated crops, but underpredicts biomass accumulation and yields if plants experience periods of water stress. Understanding this discrepancy and improving the model has been the objective of this work. Biomass underprediction has occurred because plants in the model did not accumulate enough carbon to convert into biomass. Carbon accumulation is affected by opening and closing leaf stomata. Stomata close partly or fully when the water stress occurs. The PNUTGRO model uses the widespread assumption that, if stomata become partly closed, the percentage of decrease in transpiration and carbon accumulation will be the same. We used the mechanistic leaf gas exchange model 2DLEAF to test this assumption. This model has shown that, in peanut leaves, the percentages of decrease in transpiration and carbon accumulation due to stomata closure are very different in moderate and strong stress conditions. This finding was used to correct the PNUTGRO gas exchange module, and the PNUTGRO performance grossly improved after that. Results represent a significant step towards using the peanut growth model PNUTGRO on farms as a decision support tool.
Technical Abstract: The PNUTGRO 1.02 model was used to simulate peanut (Arachis hypogaea L.) growth in Cordoba, Argentina. The model was parameterized and tested with the data from five field experiments with cv. Florman INTA conducted in three different locations. The model, calibrated to accurately predict biomass and yield in irrigated conditions, and optimized to depict atmospheric water demand and soil/plant water supply in dryland conditions, underpredicted biomass and yield in experiments with water stress. This suggests that the mechanisms of drought tolerance of the peanut crop are described in the model not adequately. The effects of water stress are accounted for by the factor SWFAC, a water supply/demand ratio, and the daily simulated photosynthesis is multiplied by this factor. This procedure assumes that stomatal control affects transpiration and photosynthesis proportionally. To improve the PNUTGRO simulation quality in water stress conditions, the relations of transpiration and photosynthesis were analyzed with the leaf-level models, 2DLEAF and the conductance analogy model. In the PNUTGRO model, the way in which water stress affects photosynthesis was modified to account for the nonlinear relation between transpiration and photosynthesis. Another factor, WSP was introduced instead of SWFAC to multiply daily photosynthesis. For an additional parameter WSFEXP an optimal value equal to 2.0 was obtained by simultaneously optimizing WSFEXP and the runoff parameter CN2, fitting biomass data from the mentioned above field experiments. This fine-tuning the model by (1) analysis of the gas exchange on the deeper level of organization (downscaling from the crop level to the leaf level), (1) using nonlinear dependencies instead of the linear ones (upscaling back