Key Laboratory of Sustainable Forest Ecosystem Management-Ministry of Education, School of Forestry, Northeast Forestry University, Harbin, 150040, Heilingjiang, People's Republic of China.
School of Forestry, Hebei Agricultural University, Baoding, 071001, Hebei, People's Republic of China.
Sci Rep. 2020 Jul 15;10(1):11664. doi: 10.1038/s41598-020-68429-w.
A photosynthetic light-response (PLR) curve is a mathematical description of a single biochemical process and has been widely applied in many eco-physiological models. To date, many PLR measurement designs have been suggested, although their differences have rarely been explored, and the most effective design has not been determined. In this study, we measured three types of PLR curves (High, Middle and Low) from planted Larix olgensis trees by setting 31 photosynthetically active radiation (PAR) gradients. More than 530 million designs with different combinations of PAR gradients from 5 to 30 measured points were conducted to fit each of the three types of PLR curves. The influence of different PLR measurement designs on the goodness of fit of the PLR curves and the accuracy of the estimated photosynthetic indicators were analysed, and the optimal design was determined. The results showed that the measurement designs with fewer PAR gradients generally resulted in worse predicted accuracy for the photosynthetic indicators. However, the accuracy increased and remained stable when more than ten measurement points were used for the PAR gradients. The mean percent error (M%E) of the estimated maximum net photosynthetic rate (P) and dark respiratory rate (R) for the designs with less than ten measurement points were, on average, 16.4 times and 20.1 times greater than those for the designs with more than ten measurement points. For a single tree, a unique PLR curve design generally reduced the accuracy of the predicted photosynthetic indicators. Thus, three optimal measurement designs were provided for the three PLR curve types, in which the root mean square error (RMSE) values reduced by an average of 8.3% and the coefficient of determination (R) values increased by 0.3%. The optimal design for the High PLR curve type should shift more towards high-intensity PAR values, which is in contrast to the optimal design for the Low PLR curve type, which should shift more towards low-intensity PAR values.
光合作用光响应（PLR）曲线是单个生化过程的数学描述，已广泛应用于许多生态生理模型。迄今为止，已经提出了许多 PLR 测量设计，尽管它们的差异很少被探索，也没有确定最有效的设计。在这项研究中，我们通过设置 31 个光合有效辐射（PAR）梯度，从种植的落叶松树上测量了三种类型的 PLR 曲线（高、中、低）。我们进行了超过 5.3 亿次设计，其中不同 PAR 梯度组合从 5 到 30 个测量点，以拟合三种类型的 PLR 曲线中的每一种。分析了不同 PLR 测量设计对 PLR 曲线拟合度和估计光合指标精度的影响，并确定了最佳设计。结果表明，PAR 梯度测量设计较少通常会导致对光合指标的预测精度较差。然而，当使用超过十个测量点进行 PAR 梯度时，精度会增加并保持稳定。对于 PAR 梯度少于十个测量点的设计，估计最大净光合速率（P）和暗呼吸速率（R）的平均百分比误差（M%E）平均分别是 PAR 梯度大于十个测量点的设计的 16.4 倍和 20.1 倍。对于单棵树，独特的 PLR 曲线设计通常会降低预测光合指标的精度。因此，为三种 PLR 曲线类型提供了三个最佳测量设计，其中根均方误差（RMSE）值平均降低了 8.3%，决定系数（R）值提高了 0.3%。高 PLR 曲线类型的最佳设计应更倾向于高强度 PAR 值，而低 PLR 曲线类型的最佳设计应更倾向于低强度 PAR 值。
