基于机器学习的滴灌玉米光合响应特征

doi:10.3864/j.issn.0578-1752.2019.17.003

摘要/Abstract

摘要：

【目的】提出一种优化模型精度的机器学习网格搜索方法,解决滴灌玉米光合响应曲线模型参数确定难、精度低等问题,为滴灌玉米光合生理机制及光合响应特征提供新思路。【方法】2017年和2018年以宁夏玉米主栽品种（TC19）为试验材料,设置6个施钾水平（0（K0）、90 kg·hm ^-2（K1）、180 kg·hm ^-2（K2）、270 kg·hm ^-2（K3）、360 kg·hm ^-2（K4）、450 kg·hm ^-2（K5））,使用Li-6400XT光合仪测定不同钾肥水平下玉米吐丝期光响应曲线。运用机器学习网格搜索法和非线性回归分析法对基于直角双曲线修正模型的光响应曲线进行拟合。选取决定系数（R ²）、均方根误差（RMSE）及平均绝对误差（MAE）对模型精度进行评价。【结果】在玉米吐丝期,叶片光合参数Pn、Tr和Gs随施钾量的增加呈先增大后减小的趋势。拟合评价结果表明,在K0和K1处理下机器学习方法计算效果优于传统方法,R ²均大于0.991,RMSE均小于1.487,MAE均小于1.350。在K2—K5处理下,2种方法拟合效果相当,R ²均大于0.993,RMSE均小于0.952、MAE均小于0.860。最优拟合方法（网格搜索法）对光响应特征参数计算结果表明,α、Pn_max、Rd、LSP和LCP的变化趋势与其光合参数相似。在施钾量为360 kg·hm ^-2（K4）时,各光响应特征参数均达到最大,在450 kg·hm ^-2（K5）时出现光抑制现象。【结论】基于机器学习的网格搜索法可准确地拟合宁夏滴灌玉米光响应特征,且施钾量为360 kg·hm ^-2时玉米光合性能达到最佳。

关键词: 玉米, 光合参数, 光响应曲线, 机器学习, 模型优化

Abstract:

【Objective】 This study proposed an optimized grid search method based on machine learning to solve the problem of model parameters of the photosynthetic light-response curve for drip-irrigated maize, which was often hard to determine and possesses low precision, so as to provide new ideas for photosynthetic characteristics and mechanisms of drip-irrigated maize in Ningxia. 【Method】 The experiment was conducted in 2017 and 2018 with the maize cultivar TC19, which was widely cultivated in Ningxia. Six levels of potassium application (0 (K0), 90 kg·hm ^-2 (K1), 180 kg·hm ^-2 (K2), 270 kg·hm ^-2 (K3), 360 kg·hm ^-2 (K4), 450 kg·hm ^-2 (K5)) were set, and the portable gas exchange system (Li-6400XT) was used to measure the light-response curves of maize under different potassium levels at silking stage. The grid search method based on machine learning and nonlinear regression analysis was used to revise the light response curve based on the right angle and hyperbolic correction model. The correlation coefficient (R ²), root-mean-square error (RMSE) and mean absolute error (MAE) were used to evaluate the accuracy of the model. 【Result】 The results showed that the photosynthetic parameters (Pn), transpiration rate (Tr) and stomatal conductance (Gs) of maize leaves increased first and then decreased with the increase of potassium application rate. The results of fitting evaluation indicated that the calculation results of machine learning method under K0 and K1 were better than the traditional method, in which R ² was greater than 0.991, RMSE and MAE were less than 1.487 and 1.350, respectively. The two methods have similar fitting effect under K2-K5, while R ²was greater than 0.993, RMSE and MAE was less than 0.952 and 0.860, respectively. The result of optical response characteristic parameter calculation by using the optimum fitting method (grid search method) showed that the trends of α, Pn_max, Rd, LSP and LCP were similar to their photosynthetic parameters. When the potassium application rate was 360 kg·hm ^-2 (K4), the light response characteristic parameters reached the maximum value, however, the light suppression phenomenon occurred at 450 kg·hm ^-2 (K5). 【Conclusion】 The grid search method based on machine learning could accurately fit the photo-response characteristics of drip-irrigated maize in Ningxia, and the photosynthetic performance of maize was the best when the potassium application rate was 360 kg·hm ^-2.

Key words: maize, photosynthetic parameters, light response curve, machine learning, model optimization

刘慧芳,贺正,贾彪,刘志,李振洲,付江鹏,慕瑞瑞,康建宏. 基于机器学习的滴灌玉米光合响应特征[J]. 中国农业科学, 2019, 52(17): 2939-2950.

LIU HuiFang,HE Zheng,JIA Biao,LIU Zhi,LI ZhenZhou,FU JiangPeng,MU RuiRui,KANG JianHong. Photosynthetic Response Characteristics of Maize Under Drip Irrigation Based on Machine Learning[J]. Scientia Agricultura Sinica, 2019, 52(17): 2939-2950.

0
/ / 推荐

导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks

链接本文: https://www.chinaagrisci.com/CN/10.3864/j.issn.0578-1752.2019.17.003

https://www.chinaagrisci.com/CN/Y2019/V52/I17/2939

图/表 8

表1

图1

图2

图3

图4

图5

表2

表3

参考文献 36

[1]	LI D F, ZHANG S Y, ZHANG G C, LIU X, XIA X X, ZHANG S S, WEI X, FANG X C . Application of five light-response models in the photosynthesis of Populus× Puramericana cv. ‘Zhonglin46’ leaves. Applied Biochemistry & Biotechnology, 2015,176(1):86-100.
[2]	CHEN Z Y, PENG Z S, YANG J, CHEN W Y, OUYANG Z M . A mathematical model for describing light-response curves in Nicotiana Tabacum L. Photosynthetica, 2011,49(3):467-471.
[3]	FAN Z L, QUAN Q M, LI Y X, JUN Y, WANG S . Exploring the best model for describing light-response curves in two Epimedium species. Technology and Health Care, 2015,23(s1):9-13.
[4]	LAROCQUE G R . Coupling a detailed photosynthetic model with foliage distribution and light attenuation functions to compute daily gross photosynthesis in sugar maple (Acer saccharum Marsh.) stands. Ecological Modelling, 2002,148(3):213-232.
[5]	STEWART D W, COSTA C, DWYER L M, SMITH D L, HAMILTON R I, MA B L . Canopy structure, light interception, and photosynthesis in maize. Agronomy Journal, 2003,95(6):1465-1474.
[6]	任佰朝, 张吉旺, 董树亭, 赵斌, 刘鹏 . 生育前期淹水对夏玉米冠层结构和光合特性的影响. 中国农业科学, 2017,50(11):2093-2103. doi: 10.3864/j.issn.0578-1752.2017.11.015
	REN B C, ZHANG J W, DONG S T, ZHAO B, LIU P . Effect of waterlogging at early period on canopy structure and photosynthetic characteristics of summer maize. Scientia Agricultura Sinica, 2017,50(11):2093-2103. (in Chinese) doi: 10.3864/j.issn.0578-1752.2017.11.015
[7]	马莉, 王全九 . 不同灌溉定额下春小麦光合光响应特征研究. 农业机械学报. 2018,49(6):271-277.
	MA L, WANG Q J . Photosynthetic and light response characteristics of spring wheat under different irrigation schedule. Transactions of the Chinese Society for Agricultural Machinery, 2018,49(6):271-277. (in Chinese)
[8]	GARDINER E S, KRAUSS K W . Photosynthetic light response of flooded cherry bark oak (Quercus pagoda) seedlings grown in two light regimes. Tree Physiology, 2001,21(15):1103-1111.
[9]	YE Z P . A new model for relationship between irradiance and the rate of photosynthesis in Oryza sativa. Photosynthetica, 2007,45(4):637-640.
[10]	YE Z P, SUGGETT D J, ROBAKOWSKI P, KANG H J . A mechanistic model for the photosynthesis-light response based on the photosynthetic electron transport of photosystem II in C₃ and C₄ species. New Phytologist, 2013,199(1):110-120.
[11]	YE Z P, ROBAKOWSKI P, SUGGETT D J . A mechanistic model for the light response of photosynthetic electron transport rate based on light harvesting properties of photosynthetic pigment molecules. Planta, 2013,237(3):837-847.
[12]	李义博, 宋贺, 周莉, 许振柱, 周广胜 . C₄植物玉米的光合-光响应曲线模拟研究. 植物生态学报. 2017,41(12):1289-1300.
	L I Y B, SONG H, ZHOU L, XU Z Z, ZHOU G S . Modeling study on photosynthetic-light response curves of a C₄ plant, maize. Chinese Journal of Plant Ecology, 2017,41(12):1289-1300. (in Chinese)
[13]	ZHAI Y F, CUI L J, ZHOU X, GAO Y, FEI T, GAO W X . Estimation of nitrogen, phosphorus, and potassium contents in the leaves of different plants using laboratory-based visible and near-infrared reflectance spectroscopy: Comparison of partial least-square regression and support vector machine regression methods. International Journal of Remote Sensing, 2013,34(7):2502-2518.
[14]	王丽爱, 马昌, 周旭东, 訾妍, 朱新开, 郭文善 . 基于随机森林回归算法的小麦叶片SPAD值遥感估算. 农业机械学报, 2015,46(1):259-265.
	WANG L A, MA C, ZHOU X D, ZI Y, ZHU X K, GUO W S . Estimation of wheat leaf SPAD value using RF algorithmic model and remote sensing data. Transactions of the Chinese Society for Agricultural Machinery, 2015,46(1):259-265. (in Chinese)
[15]	依尔夏提·阿不来提, 买买提·沙吾提, 白灯莎·买买提艾力, 安申群, 马春玥 . 基于随机森林法的棉花叶片叶绿素含量估算. 作物学报, 2019,45(1):81-90.
	YIERXIATI A B L T, MAIMAITI S W T, BAIDENGSHA M M T A L, AN S Q, MA C Y . Estimation of leaf chlorophyll content in cotton based on the Random Forest approach. Acta Agronomica Sinica, 2019,45(1):81-90. (in Chinese)
[16]	曾继业, 谭正洪, 三枝信子 . 近似贝叶斯法在光合模型参数估计中的应用. 植物生态学报, 2017,41(3):378-385. doi: 10.17521/cjpe.2016.0067
	ZENG J Y, TAN Z H, SANZHI X Z . Using approximate Bayesian computation to infer photosynthesis model parameters. Chinese Journal of Plant Ecology, 2017,41(3):378-385. (in Chinese) doi: 10.17521/cjpe.2016.0067
[17]	SINGH S K, REDDY V R . Co-regulation of photosynthetic processes under potassium deficiency across CO₂ levels in soybean: Mechanisms of limitations and adaptations. Photosynthesis Research, 2018,137(2):183-200.
[18]	KANAI S, OHKURA K, ADU GYAMFI J J, MOHAPATRA P K, NGUYEN N T, SANEOKA H, FUJITA K . Depression of sink activity precedes the inhibition of biomass production in tomato plants subjected to potassium deficiency stress. Journal of Experimental Botany, 2007,58(11):2917-2928.
[19]	QU C X, LIU C, ZE Y G, GONG X L, HONG M M . Inhibition of nitrogen and photosynthetic carbon assimilation of maize seedlings by exposure to a combination of salt stress and potassium-deficient stress. Biological Trace Element Research, 2011,144(1/3):1159-1174.
[20]	WANG X G, ZHAO X H, JIANG C J, LI C H, CONG S, WU D, CHEN Y Q, YU H Q, WANG C Y . Effects of potassium deficiency on photosynthesis and photoprotection mechanisms in soybean (Glycine max(L.)Merr.). Journal of Integrative Agriculture, 2015,14(5):856-863.
[21]	杜琪, 王宁, 赵新华, 沙德剑, 张艳正, 赵凯能, 党现什, 于海秋 . 低钾胁迫对玉米苗期光合特性和光系统Ⅱ性能的影响. 核农学报, 2019,33(3):592-599.
	DU Q, WANG N, ZHAO X H, SHA D J, ZHANG Y Z, ZHAO K N, DANG X S, YU H Q . Effects of potassium deficiency on photosynthesis and performance of photosystem Ⅱ in maize seedling stage. Journal of Nuclear Agricultural Sciences, 2019,33(3):592-599. (in Chinese)
[22]	ZHAO D L, OOSTERHUIS D M, BEDNARZ C W . Influence of potassium deficiency on photosynthesis, chlorophyll content, and chloroplast ultrastructure of cotton plants. Photosynthetica, 2001,39(1):103-109.
[23]	ZHAO X H, DU Q, ZHAO Y, WANG H J, LI Y J, WANG X G, YU H Q . Effects of different potassium stress on leaf photosynthesis and chlorophyll fluorescence in maize at seedling stage. Agricultural Sciences, 2016,7(1):44-53.
[24]	夏乐, 于海秋, 郭焕茹, 赵尚文, 姚晓旭, 曹敏建 . 低钾胁迫对玉米光合特性及叶绿素荧光特性的影响. 玉米科学, 2008,16(6):71-74.
	XIA L, YU H Q, GUO H R, ZHAO S W, YAO X X, CAO M J . Effects of potassium deficiency on photosynthetic characters and chlorophyll fluorescence characters in maize plants. Journal of Maize Sciences, 2008,16(6):71-74. (in Chinese)
[25]	张兴风, 刘泽人, 黄兴法, 杨建国, 李光永 . 宁夏膜下滴灌玉米不同施肥模式的试验研究. 节水灌溉, 2016,8:57-60.
	ZHANG X F, LIU Z R, HUANG X F, YANG J G, LI G Y . Experimental study on different fertilization patterns at different growth stages of maize under mulched drip irrigation in Ningxia. Water Saving Irrigation. 2016,8:57-60. (in Chinese)
[26]	刘学军, 翟汝伟, 李真朴, 刘平, 王文 . 宁夏扬黄灌区玉米滴灌水肥一体化灌溉施肥制度试验研究. 中国农村水利水电, 2018,431(9):74-78.
	LIU X J, ZHAI R W, LI Z P, LIU P, WANG W . Experimental research on the application of integrated fertilizer and fertilizer system for maize drip irrigation in Ningxia Yellow River irrigation area. China Rural Water and Hydropower, 2018,431(9):74-78. (in Chinese)
[27]	李哲, 屈忠义, 任中生, 杨少东, 续喆, 贾咏林, 胡敏 . 河套灌区膜下滴灌高频施肥促进玉米生长及产量研究. 节水灌溉, 2018,278(10):1-4.
	LI Z, QU Z Y, REN Z S, YANG S D, XU Z, JIA Y L, HU M . A study on the promotion of drip irrigation under plastic film with high frequency fertigation for maize growth and yield in Hetao irrigation district. Water Saving Irrigation, 2018,278(10):1-4. (in Chinese)
[28]	SHANKAR A, SINGH A, KANWAR P, SRIVASATAVA A K, PANDEY A, SUPRASANNA P, KAPOOR S, PANDEY G K . Gene expression analysis of rice seedling under potassium deprivation reveals major changes in metabolism and signaling components. PLoS ONE, 2013,8(7):e70321.
[29]	ALI S, HAFEEZ A, MA X L, TUNG S A, LIU A, SHAH A N, CHATTHA M S, ZHANG Z, YANG G Z . Potassium relative ratio to nitrogen considerably favors carbon metabolism in late-planted cotton at high planting density. Field Crops Research, 2018,223:48-56.
[30]	TRANKNER M, TAVAKOL E, JÁKLI B . Functioning of potassium and magnesium in photosynthesis, photosynthate translocation and photoprotection. Physiologia Plantarum, 2018,163(3):414-431.
[31]	HAEDER H E, MENGEL K, FORSTER H . The effect of potassium on translocation of photosynthates and yield pattern of potato plants. Journal of the Science of Food & Agriculture, 2010,24(12):1479-1487.
[32]	叶子飘, 张海利, 黄宗安, 杨小龙, 康华靖 . 叶片光能利用效率和水分利用效率对光响应的模型构建. 植物生理学报, 2017,53(6):226-232.
	YE Z P, ZHANG H L, HUANG Z A, YANG X L, KANG H J . Model construction of light use efficiency and water use efficiency based on a photosynthetic mechanistic model of light response. Plant Physiology Journal, 2017,53(6):226-232. (in Chinese)
[33]	MIAO Z W, XU M, RICHARD G J R L, WNAG Y F . Comparison of the A-Cc curve fitting methods in determining maximum ribulose 1.5-bisphosphate carboxylase/oxygenase carboxylation rate, potential light saturated electron transport rate and leaf dark respiration. Plant Cell & Environment, 2009,32(2):109-122.
[34]	SHARP R E, MATTHEWS M A, BOYER J S . Kok effect and the quantum yield of photosynthesis light partially inhibits dark respiration. Plant Physiology, 1984,75(1):95-101.
[35]	叶子飘 . 光合作用对光和CO₂响应模型的研究进展. 植物生态学报, 2010,34(6):727-740. doi: 10.3773/j.issn.1005-264x.2010.06.012
	YE Z P . A review on modeling of responses of photosynthesis to light and CO₂. Chinese Journal of Plant Ecology, 2010,34(6):727-740. (in Chinese) doi: 10.3773/j.issn.1005-264x.2010.06.012
[36]	赵丽, 贺玉晓, 魏雅丽, 刘刚才 . 干热河谷紫色土区不同复合肥施肥量对玉米苗期光响应特性的影响. 干旱地区农业研究, 2018,36(1):10-18.
	ZHAO L, HE Y X, WEI Y L, LIU G C . Light response characteristics of maize seedling under different amount of compound fertilizer in the purple soil at dry-hot valley. Agricultural Research in the Arid Areas, 2018,36(1):10-18. (in Chinese)

试验地 Test location	年份 Year	pH	有机质 OM (g·kg^-1)	全氮 Total N (g·kg^-1)	全磷 Total P (g·kg^-1)	碱解氮 Available N (mg·kg^-1)	速效磷 Available P (mg·kg^-1)	速效钾 Available K (mg·kg^-1)
平吉堡农场 Pingjipu farm	2017	7.92	11.51	0.82	0.59	37.42	19.11	102.48
平吉堡农场 Pingjipu farm	2018	7.57	12.78	0.70	0.45	36.64	17.42	95.34

年份 Year	方法 Method	处理 Treatment	RMSE	MAE	R²
2017	非线性回归分析法 Nonlinear regression analysis	K0	2.030	1.869	0.990
		K1	1.319	1.079	0.992
		K2	1.147	0.899	0.993
		K3	0.828	0.656	0.995
		K4	0.610	0.498	0.994
		K5	0.446	0.379	0.995
	网格搜索法 Grid search method	K0	1.335	1.096	0.993
		K1	1.312	1.092	0.993
		K2	0.952	0.622	0.994
		K3	0.708	0.617	0.995
		K4	0.546	0.504	0.996
		K5	0.662	0.557	0.995
2018	非线性回归分析法 Nonlinear regression analysis	K0	1.827	1.445	0.987
		K1	1.731	1.323	0.991
		K2	0.872	0.733	0.993
		K3	0.858	0.717	0.994
		K4	0.644	0.566	0.996
		K5	0.312	0.226	0.997
	网格搜索法 Grid search method	K0	1.487	1.350	0.991
		K1	0.975	0.860	0.993
		K2	0.839	0.716	0.993
		K3	0.774	0.620	0.995
		K4	0.750	0.637	0.996
		K5	0.330	0.315	0.996

年份 Year	处理 Treatment	表观量子效率 Q (μmol·μmol^-1)	最大光合速率 Pn_max (μmol·m^-2·s^-1)	光补偿点 Ic (μmol·m^-2·s^-1)	光饱和点 Is (μmol·m^-2·s^-1)	暗呼吸速率 Rd (μmol·m^-2·s^-1)	决定系数 R²
2017	K0	0.041	23.959	55.922	1453.189	2.161	0.987
	K1	0.046	27.239	59.061	1627.178	2.510	0.991
	K2	0.050	30.672	68.087	1965.095	3.260	0.993
	K3	0.053	32.953	73.295	2126.701	3.456	0.994
	K4	0.061	34.653	76.505	2694.469	4.635	0.996
	K5	0.056	33.997	63.495	2034.091	4.066	0.997
2018	K0	0.042	25.368	56.110	1504.405	2.364	0.99
	K1	0.043	31.341	61.278	1568.254	2.622	0.992
	K2	0.047	32.807	63.761	1984.520	3.758	0.993
	K3	0.056	34.407	67.300	2282.035	3.982	0.995
	K4	0.063	35.678	78.092	2893.226	5.00	0.994
	K5	0.058	35.256	71.462	2462.256	4.228	0.995