Effects of Functional Organic Materials on Fresh Weight,Quality,Single-melon Weight and SPAD Value of Watermelon

[Objectives]This study was conducted to explore a functional organic material formula suitable for watermelon cultivation with high quality,high yield and high efficiency.[Methods]Four treatments were set in the experiment,namely four functional organic materials,garlic straw treatment(T_(1)),onion straw treatment(T_(2)),garlic straw+sheep manure treatment(T_(3))and onion straw+chicken manure treatment(T_(4)),to investigate the effects of different functional organic materials on fresh weight,quality,single-melon weight and SPAD value of watermelon.[Results]The effects of different functional organic materials on fresh weight,quality,single-melon weight and SPAD value of watermelon were quite different.The fresh weight,quality,single-melon weight and SPAD value of watermelon were higher in treatment T_(3)applying garlic straw and sheep manure and treatment T_(4)applying onion straw and chicken manure than in treatment T_(1)applying garlic straw and treatment T_(2)applying onion straw.Specifically,the fresh weight of whole plant was the highest in treatment T_(3),followed by treatment T_(4),and the values of the two treatments increased by 12.83%and 5.94%respectively compared with treatment T_(1);the weight of single melon was the highest in treatment T_(3),followed by treatment T_(4),and the values of the two treatments increased by 42.45%and 31.77%respectively compared with treatment T_(2);and the SPAD values of treatments T_(3)and T_(4)were significantly higher than those of treatments T_(1)and T_(2),and the value of treatment T_(3)was the largest.[Conclusions]This study provides theoretical support for the popularization and application of fertilization techniques combining organic fertilizers and reduced chemical fertilizers for watermelon.

Correlation Analysis of SPAD Value with Chlorophyll Content and Economic Yield Traits of Brassica napus L.

[Objective] The aim was to compare differences of SPAD value, chloro- phyll content, agronomic characters, economic characters and yield traits to analyze correlation of SPAD value with other indices and establish regression functions. [Method] Based on 34 Brassica napus L. varieties, SPAD value, chlorophyll content, agronomic characters, economic characters and yield traits were measured and re- gression functions were established according to correlations. [Result] SPAD value, chlorophyll content, agronomic and economic characters and yield traits all achieved significant level in differences among 34 varieties. Specifically, SPAD value was of extremely significant correlation with chlorophyll a and b, total chlorophyll and carotenoid, and the correlation from high to low was chl-b〉chl-z〉chl-a〉chl-x. SPAD value was of significantly positive correlation with total pod number per plant, plant height, seed number per pod, yield per plant and harvest yield, and of insignificant correlation with branch point height, effective branch number, pod density of main stem, and pod length. [Conclusion] It is simple and rapid to predict chlorophyll con- tent, economic characters and yields of Brassica napus L. with SPAD value and re- gression functions.

Effects of Nitrogen Level and High Temperature Stress on Yield, SPAD Value and Soluble Sugar Content of Early Rice Ganxin 203

This study aimed to investigate the combined effects of nitrogen level （high and ordinary） and high temperature stress （37 ℃, 4 d） at the late panicle ini- tiation stage on yield, SPAD value and soluble sugar content of Ganxin 203, an early rice cultivar. The results showed that under both high and ordinary nitrogen levels, high temperature stress reduced the seed-setting rate, yield per stem, SPAD value and soluble sugar content of Ganxin 203; under both high and ordinary tem- peratures, high nitrogen level increased the seed-setting rate, yield per stem and soluble sugar content of Ganxin 203; the seed-setting rate, yield per stem and yield per plant of Ganxin 203 under high temperature and high nitrogen level were higher than those under high temperature and ordinary nitrogen level. It suggests that ap- propriate high nitrogen level contributes to weakening the negative effects of high temperature stress on rice yield.

Effects of SPAD Decline Value of Mid-season Hybrid Rice Leaves after Full-heading Stage on Productivity of Rice Fertilized with Nitrogen

[Objective] The aim was to research relationship between SPAD decline index after full-heading stage （SDIFHS） and productivity of rice. fertilized with nitro- gen in order to provide theoretical and practical references for selection and breed- ing of rice varieties. [Method] From 2008 to 2010, 18 mid-season hybrid rice vari- eties were researched every year to explore relationship＇ between SDIFHS and pro- ductivity of rice fertilized with nitrogen. [Result] The productivity of rice fertilized with nitrogen was of extremely significant positive corretation with SDIFHS, because the higher SPAD decline index is, the higher LAI decline index and the transformation ratio of dry matter to spikes in overground plant would be. [Conclusion] The re- search established a new method to predict productivity of rice fertilized with nitro- gen based on SPAD decline index.

华南晚籼杂交稻叶片SPAD值及其一般配合力和遗传效应的动态变化

本研究旨在揭示华南晚籼杂交稻叶片SPAD值一般配合力及遗传参数的动态变化规律,为华南杂交水稻的高产高效新品种培育提供相关的科学依据。以华南地区生产上广泛应用的3个籼稻不育系和6个籼稻恢复系配置了18个不完全双列杂交组合,分析不同发育阶段叶片SPAD值及其一般配合力及遗传效应的动态变化规律。结果表明:杂种及其亲本叶片SPAD值移栽后呈逐渐下降趋势,于幼穗分化期(栽后43或50 d)达到最低点,始穗期(60 DAT)后快速下降。不同组合以及同一组合在不同发育时期叶片SPAD值存在较大差异。灌浆结实期至蜡熟期,‘荣丰A’、‘五丰A’、‘华占’和‘广恢308’及其相应组合叶片SPAD值下降较快,而‘明恢63’和‘桂99’及其相应杂交组合依然保持相对较高的SPAD值,并与亲本SPAD值一般配合力的动态变化相一致。除分蘖始期叶片SPAD值以特殊配合力为主外,达70.76%,其余发育阶段叶片SPAD值均以一般配合力起主导作用。杂种叶片SPAD值在生殖生长阶段的遗传力(21.90%~63.89%),显著高于其营养生长阶段的遗传力(8.02%~14.79%);其遗传效应以加性效应为主,同时存在显性或/和上位性效应,始穗期至抽穗扬花期的显性或/和上位性效应增大,达18.37%和22.02%。发育前期和中期,叶片SPAD值受环境条件影响较小,后期则较大地受到环境条件影响。

基于无人机多光谱影像的春小麦SPAD值估测研究

【目的】探究无人机遥感技术对春小麦SPAD值的无损估测,建立小田块尺度上的春小麦SPAD值估测模型。【方法】利用多光谱无人机获取2022年内蒙古河套地区春小麦冠层多光谱数据,并提取5种植被指数,通过分析植被指数与植株SPAD值的相关性,结合支持向量机回归(SVR)、偏最小二乘回归(PLSR)和随机森林回归(RFR)构建春小麦关键生育时期SPAD值估测模型,并筛选各时期SPAD值最优估测模型。【结果】通过与5种植被指数配对来确定最佳模型,在春小麦的拔节期、抽穗期、开花期和灌浆期,这15种组合方法均可对SPAD值进行估测。对3种建模方法的最优模型进行比较发现,NDVI+RFR模型在估测春小麦SPAD值方面具有最高的准确性和良好的鲁棒性,其R^(2)=0.96,RMSE=0.19。【结论】利用无人机估测春小麦的SPAD值是可行的,最优模型为NDVI+RFR模型。

基于SPAD值的不同海拔烤烟采收成熟度研究

为探索不同海拔下烟叶成熟度采收标准及采收时间,以张家界市主栽品种K326为研究对象,使用SPAD-502叶绿素仪测定了400和600 m海拔区域下上、中、下各部位不同成熟度鲜烟叶的SPAD值,分析不同成熟度下烟叶外观质量、感官质量及经济性状,找出不同海拔下各部位烟叶适宜采收的SPAD值。结果表明:400 m海拔下部烟叶适宜采收的SPAD值为26.57~29.99,中部鲜烟叶SPAD值为15.48~23.08,上部鲜烟叶SPAD值为17.19~20.81;600 m海拔下部烟叶适宜采收的SPAD值为21.60~24.06,中部鲜烟叶SPAD值为19.82~24.40,上部鲜烟叶SPAD值为13.49~16.65,烤后烟叶外观质量表现较优,感官质量属于较高档次。

基于杂交水稻叶片SPAD值的氮素营养诊断模型的初步构建

为探明水稻叶片SPAD值的最佳测定叶位和最佳的SPAD值次级指标,并构建基于水稻叶片SPAD值的氮素营养诊断模型,开展以品种为主区,氮肥施用量为副区的双因素裂区设计试验。参试品种为Q优6号和宜香优2115,氮肥施用量设5个水平(纯N 0、75、150、225、300 kg/hm^(2)),分析叶片的敏感性、代表性和稳定性,并探讨SPAD值次级指标与施氮量和叶片含氮量之间,及叶片氮积累量与叶片含氮量和产量之间的关系。结果表明,叶片敏感性、代表性和稳定性大小顺序分别为L4(顶4叶)>L3(顶3叶)>L1(顶1叶)>L2(顶2叶)和L3>L4>L2>L1、L2>L3>L4>L1,可见,L3、L4可作为氮素营养诊断的共同理想指示叶。选择L3、L4的SPAD值几何平均数(GMSI34)作为最佳的SPAD值次级指标。由叶片氮积累量与产量和叶片含氮量的抛物线方程、一元线性回归方程,GMSI34与叶片含氮量的指数方程,求得拔节期、孕穗期、抽穗期的SPAD值次级指标的临界值(GMSI34_(临))分别为48.5、44.3和42.9,与实际获取的GMSI34实相比较,若GMSI34_(实)-GMSI34_(临)<0时,则表明水稻缺氮,需要追氮,若GMSI34_(实)-GMSI34_(临)≥0,则表明水稻氮营养充足,无需追氮。综上,基于杂交水稻叶片SPAD值的氮素营养诊断模型,可为杂交水稻高效氮肥管理提供技术支持。

基于叶片SPAD值的微型盆栽月季氮素营养诊断研究

氮素是微型盆栽月季生长发育的必需元素,实时监测植株生长过程中氮素的变化,精准估测氮素含量尤为重要。设置4个施氮水平(包含高氮处理),分析施氮量对微型盆栽月季生长发育的影响,测定不同施氮水平下4种微型盆栽月季在不同生长时期、不同叶片层位的SPAD值,利用线性及非线性回归分析等方法,构建基于叶片SPAD值的单一品种和混合品种叶片氮含量估测模型,并对此模型进行检验和通用性评价,利用SPAD氮饱和指数评估临界氮含量。结果显示,4个不同微型月季品种在整个生长发育时期的营养生长趋势相似,且在不同施氮水平下性状指标具有差异。随着施氮量的增加,叶片氮含量和叶片SPAD值均呈先上升后平稳的趋势。叶片SPAD值与叶片氮含量之间存在显著相关性,且不同层位的叶片SPAD值与叶片氮含量的相关性存在差异。多品种混合构建的综合模型在通用性上优于单一品种模型,S1和S2时期的下层叶片、S3和S4时期的上层叶片SPAD值与叶片氮含量回归决定系数较高,构建的模型分别为y=0.0009x^(2)+0.6115x+0.8047、y=0.002x2+0.7118x+2.3382、y=0.1279x^(1.4938)、y=-0.0015x^(2)+0.5204x+3.7738。氮饱和指数利用多项式模型估测较为可靠。综上所述,施氮水平、生育期、叶片层位均会影响SPAD值与叶片氮含量的相关性,可利用混合品种建立的叶片氮含量估测综合模型作为通用模型,利用氮饱和指数多项式方程进行氮素诊断较为准确,为微型盆栽月季氮素营养实时诊断提供理论数据。

田间即时鲜烟叶SPAD值预测和成熟度识别方法

成熟程度判定作为烟叶采收时的重要工作,需要满足即时性、科学性和准确性要求。目前烟叶成熟度识别实施存在测量仪器价格昂贵且操作繁琐,无法在田间推广使用;手机摄像头的内嵌图像处理算法干扰图像有效特征,田间复杂天气环境影响图像采集一致性;现存识别算法忽略植物学领域信息,影响模型准确性和普适性等问题。据此提出一种低成本且有效的田间即时鲜烟叶SPAD值预测和成熟度识别方法,通过提高识别准确率保障烟叶后续调制质量。研发便携式拍摄装置,实现田间高质量图片采集;并以CX-26烟叶品种为研究对象,提出一种适配于田间烟叶图像的分割方法,通过提取图像目标区域特征数据,利用XGBoost算法依次搭建鲜烟叶SPAD值预测模型和成熟度识别模型。提出两个模型的集成思路,集成模型能够利用SPAD值和成熟度的强相关关系,通过预测SPAD值提高成熟度识别准确性。该方法在各项评价指标中均表现优秀,其中SPAD值预测平均误差为0.4703,成熟度识别宏F1-Score为95.27%。研发手机APP完成拍摄装置和云端模型之间烟叶图像和预测结果的传输,实现在田间对烟叶成熟程度快速、客观、准确的即时预测。该成果可为田间农作物精准采收提供有效技术支持。

Diagnosis of Premature Senescence of Cotton Using SPAD Value

Field experiments were conducted in 2011 on an experimental farm at the Cotton Research Insti-tute, Chinese Academy of Agricultural Sciences, Anyang, China. We conducted experiments with a “SPAD-502” meter to quickly measure the relative value of chlorophyll content in the leaf blades of field cotton. Our goal was to seek a suitable leaf on a cotton plant to diagnose senescence status of crop plants at later stages of growth. We began by studying the dynamic change rule of the dis-tal-most four leaves of the cotton plant during the entire growth period with two early-maturing cultivars (CCRI 36, CCRI 50) and two mid-maturing cultivars (CCRI 41, SCRC 28). We also examined the effect of different nitrogen and potassium fertilizer rates on SPAD values of the leaves of SCRC 28. Our results suggest that SPAD values of the 1st distal stem leaves from two early cultivars can act as good indicators of senescence status in the plants, if they are measured between 10 d before the boll-opening stage and 10 d after boll opening stage. Differences of SPAD values of the 3rd distal stem leaves of two mid-maturing cultivars measured between about 15 d before the boll opening stage and 15 d after the boll opening stage can also be used to measure senescence status in these cotton cultivars. The conclusion can be used for cotton producer to manage N fertilizer better at later growth stage.

Response of Chlorophyll a, SPAD Values and Chlorophyll Fluorescence Parameters in Leaves of Apricot Affected Some Abiotic Factors

Experimental trail conducted in Jasenica, Bosnia and Herzegovina have showed apricot （Prunus persica L） leaf injures especially on the edges in the middle of the season. Leaf edges show chlorosis starting from the tip due to lack of Chlorophyll. Affected leaves start falling and renew leaves masse as respond to stress conditions which also delay fruits maturation. Technological development in detecting the stress of high plants trough the leaf fluorescence and Chlorophyll concentration have led to the introduction of instruments which utilize fluorescence as a basis for determining stress level. The paper was aimed to show whether the productivity values calculated from Chlorophyll fluorescence Fv/Fm were comparable to those produced by spectrophotometric method in detecting Chlorophyll a concentration and SPAD values detected by Chlorophyll meter （SPAD-502）. Measurements of nitrogen content were also provided in order to obtain accuracy of SPAD meter since Chlorophyll molecules contain nitrogen in their structure. To test the relationship between these three technique regression relationship was obtained. Regression coefficient between Fv/Fm values and Chlorophyll a concentration mg g-t was high. Also, strong regression coefficient was observed between Chlorophyll a and SPAD values indicating a good accuracy of this device which was also confirmed by good regression between nitrogen content and SPAD values. Fluorometer measurements on injured leaves samples of apricot have also reviled the Fv/Fm values below 0.83 which is in according with numerous authors who considered a line for indicating stress factors.

基于SPAD的西瓜氮素营养诊断与推荐施肥技术研究

为解决西瓜生产中存在盲目或过量氮肥施用导致的果实品质下降和氮素利用率低下等突出问题,开展了基于叶绿素仪测定值(SPAD值)的西瓜全生育期氮素营养诊断与推荐施肥技术研究,以期为利用SPAD仪建立田间西瓜氮素营养实时诊断技术、开发西瓜精准氮素养分管理模式提供理论依据。试验于2020-2021年在甘肃河西灌区的张掖市以‘金城5号’西瓜为材料,设置了0、80、160、240、320、400 kg/hm^(2)6个氮肥水平,通过研究不同施氮水平下西瓜关键生育时期不同叶位及叶片不同位置上的SPAD值、叶绿素含量、全氮含量的变化特征及其与西瓜产量、品质指标的相互关系,建立基于SPAD值进行西瓜氮素营养诊断与推荐施肥的实时氮肥管理(RTNM)模式,并开展RTNM模式的生产性验证。西瓜苗期顶一叶叶尖、伸蔓期顶三叶叶中、膨果期功能叶叶中位置SPAD值与叶片叶绿素、全氮含量相关性最佳,且变异系数小,是较为理想的指示叶或参照叶;通过指示叶SPAD值与西瓜产量、品质指标及施氮量的回归分析,获得西瓜产量超过60000 kg/hm^(2)的SPAD阈值范围为苗期49.1~51.0、伸蔓期60.3~64.1、膨果期64.8~67.5;西瓜果实可溶性固形物含量超过11.5%的SPAD阈值范围为苗期49.0~50.5、伸蔓期59.8~63.6、膨果期64.0~66.5;SPAD值每变动1个单位的氮肥施用量分别为苗期30.96 kg/hm^(2)、伸蔓期31.95 kg/hm^(2)、膨果期50.51 kg/hm^(2)。通过与传统施肥模式(N 240 kg/hm^(2))相比较,RTNM模式在氮肥施用量减少15.29%的条件下,西瓜产量较高、品质较优,且氮肥利用率显著提高了30.53%。综合分析得出,甘肃河西灌区‘金城5号’西瓜膜下滴灌栽培模式下,应用SPAD仪进行氮素营养诊断与推荐施肥,适宜的指示叶及SPAD阈值范围为苗期顶一叶叶尖49.1~50.5、伸蔓期顶三叶叶中60.3~63.6、膨果期功能叶叶中64.8~66.5。

基于无人机多光谱影像的水稻冠层SPAD值预测研究

比较筛选水稻冠层SPAD值估测模型,为无人机多光谱遥感反演水稻SPAD值提供依据。利用无人机获取水稻拔节期、抽穗期、乳熟期的冠层多光谱影像,选取7种常用的植被指数,利用3种回归方法建立基于植被指数的水稻叶片SPAD值反演模型。结果表明,在不同生育期与水稻叶片SPAD值相关系数最高的植被指数不相同,拔节期最高的是GNDVI,抽穗期最高的是CIGreen,乳熟期最高的是CIRededge。抽穗期是水稻叶片SPAD值反演的最佳时期,模型具有较好的建模精度和估测效果,其中多元线性回归的建模精度较高,偏最小二乘回归模型的估测效果最好。试验结果可为水稻长势的实时无损监测提供参考。

基于无人机多光谱影像的棉花SPAD值及叶片含水量估测

利用多光谱快速准确获取棉花花铃期旱胁迫性状对棉花后期产量和纤维品质具有重要帮助。为实现棉花关键抗旱指标的快速检测,该研究以253份棉花品种为材料,设置干旱胁迫和正常灌溉2个处理,在花铃期通过大疆精灵4多光谱无人机获得图像,分析花铃期干旱胁迫和正常处理棉花的光谱反射率,结合地面调查叶绿素相对值(S_(PAD))和叶片含水量(L_(WC))数据,结合神经网络算法中的径向基函数进行模型的预测。结果显示:棉花干旱胁迫前后除红光反射值略上升,其余4个光谱均小于正常处理;通过径向基函数模型估测S_(PAD)和L_(WC)的最佳估测模型都为二次函数,决定系数R^(2)分别为0.8488和0.9366,均方根误差R_(MSE)分别为2.005和0.930,相对误差R_(E)分别为0.004和0.011;将2个模型应用于试验区影像,对S_(PAD)及L_(WC)预测值进行聚类分析,根据S_(PAD)聚类结果,试验棉花旱情等级划分为特旱、重旱、中旱、轻旱、无旱5类,根据LWC聚类结果,试验棉花旱情等级划分为特旱、重旱、中旱、干旱、轻旱、无旱6类。以上模型及分类效果良好且符合棉花干旱胁迫的特点,研究结果可为快速评价棉花抗旱性、挖掘棉花抗旱基因提供技术支持。

寒地水稻叶片SPAD值与稻米蛋白质含量的关系

为揭示寒地水稻叶片SPAD值及其衍生值与稻米蛋白质含量的关系,于2020—2021年在水稻拔节期(T1)、孕穗期(T2)、齐穗期(T3)测定顶部3片叶的SPAD值,依据盆栽试验(试验1和试验2)的数据资料建立SPAD值衍生指标与稻米蛋白质含量之间的关系模型,利用大田试验(试验3)数据资料对建立的模型进行验证。结果表明,2020年氮肥试验中A8水平(氮素施用量362.07 kg·hm^(-2))稻米蛋白质含量较A1~A7水平(A1~A7氮素施用量分别为0、51.72、103.45、155.17、206.90、258.62、310.35 kg·hm^(-2))分别极显著增加34.55%、27.44%、26.39%、22.19%、18.07%、14.39%、12.23%,而A8水平食味值较A1~A7水平分别极显著降低8.10%、5.06%、4.99%、4.10%、3.45%、2.96%、2.28%,2021年蛋白质含量、食味值变化趋势与前者相同。两年品种试验6个品种稻米蛋白质含量比较中,C6品种(三江6号)蛋白质含量较C5(龙粳21)、C4(垦粳8号)、C3(龙稻203)、C2(松粳16)、C1(松粳22)分别极显著提高2.99%、12.23%、10.43%、5.04%、15.63%,C6食味值较C5、C4、C3、C2、C1分别极显著降低1.17%、12.09%、3.54%、2.89%、7.93%。品种差异及施氮量对不同生育时期水稻顶部3片叶的SPAD值分布规律有较大影响,但水稻冠层叶片出现的两次“黑黄交替”现象不受品种的影响,其中单片叶的SPAD值受品种差异的影响,与蛋白质含量不存在相关性,借助指标SPAD_((L1+L2+L3)/3)、SPAD_(L2×L3/mean)可有效降低品种及环境差异对蛋白质含量预测结果的影响,指标SPAD_((L1+L2+L3)/3)与蛋白质含量在T1~T3期(拔节期、孕穗期、齐穗期)的拟合方程分别为:Y=0.24 X-1.94,R^(2)为0.75^(**)、Y=0.25 X-1.69,R^(2)为0.74^(**)、Y=0.27 X-2.45,R^(2)为0.72^(**);SPAD_(L2×L3/mean)拟合方程分别为:Y=0.22 X-1.05,R^(2)为0.75^(**)、Y=0.27 X-2.43,R^(2)为0.72^(**)、Y=0.26 X-2.24,R^(2)为0.72^(**)(Y为蛋白质含量,X为SPAD衍生值),拟合方程均达到极显著水平。稻米蛋白质含量和食味值评分变现为线性负相关,回归方程为Y=-4.21 X+113.32(Y为食味值,X为蛋白质含量),拟合优度R^(2)=0.93^(**),达到极显著水平。综上,借助指标SPAD_((L1+L2+L3)/3)和SPAD_(L2×L3/mean)能够实现快捷、无损和实时预测稻米蛋白质含量,在一定程度上判定出稻米蒸煮食味品质的优劣,达到按质收获以及对品质实时监测的要求,促进优质寒地水稻的可持续发展。

不同生育期小麦冠层SPAD值无人机多光谱遥感估算

随着长江中下游稻麦轮作区水稻成熟期的推迟,冬小麦播期的推迟已经成为影响产量的主要障碍,因此在迟播小麦中筛选抗性较好的品种很有必要。该研究旨在监测冬小麦生长早期冠层叶片的相对叶绿素含量,用于迟播冬小麦品种筛选。为探讨利用无人机多光谱影像监测冬小麦叶绿素含量的可行性,基于多光谱无人机获取的5个单波段光谱反射率和15个植被指数作为自变量,经过递归特征消除法(RFE)特征变量筛选,去除冗余变量,利用后向神经网络(BP)回归算法构建冬小麦相对叶绿素含量(SPAD)值遥感反演模型。根据2020年—2021年江苏省扬州市广陵区实验点冬小麦越冬期、拔节期两个生育期的实测叶片SPAD值,结合同步获取的多光谱无人机影像,分析了这两个生育期遥感变量和SPAD值之间的相关性。并结合遥感变量之间的特征重要性排序进行特征变量筛选,筛选出的变量作为模型的输入,构建并筛选出各生育期最佳的反演模型。比较岭回归(Ridge)和梯度提升树(GBD)算法,以R^(2)和RMSE作为模型评价指标,在验证集上分析了各生育期3种模型的自学习能力和泛化能力。结果表明,经过了最优光谱信息筛选而建立的BP神经网络模型在此两个生育期的数据集上均表现出了最强的回归预测能力。R^(2)和RMSE在越冬期分别为0.806和1.861,拔节期分别为0.827和0.507。通过对无人机多光谱数据进行变量筛选,构建的优选模型BP神经网络具有较高估算精度,且表明在冬小麦的早期监测中,拔节期比越冬期效果好。利用无人机多光谱在估算迟播冬小麦SPAD值进行品种抗性筛选的方法是有价值的。

基于不同分析模型的大豆叶片SPAD值和LAI光谱估算比较

为探讨在大豆鼓粒期采用光谱技术估算叶片SPAD值和LAI的有效分析模型和方法,本研究以大田鼓粒期大豆为试验材料,在3个不同时段9:45~10:15(10AM)、11:45~12:15(12PM)和13:45~14:15(2PM)测量冠层全波段光谱反射率,并分别使用极限学习机(ELM)、偏最小二乘回归(PLSR)、支持向量机(SVM)和随机森林(RF)构建大豆叶片SPAD值和LAI估算模型,并对比不同模型分析结果的估算精度。结果表明:在各模型中,12PM和2PM测定的光谱反射率与大豆叶片SPAD值和LAI的拟合精度均高于10AM。基于RF的大豆叶片SPAD值估算模型验证集的R2为0.910,RMSE为2.006,MRE为3.684;基于RF的大豆LAI估算模型验证集的R^(2)为0.916,RMSE为0.209,MRE为4.383,与ELM、PLSR和SVM相比,有更高的估算精度。综上结果说明大豆鼓粒期在11:45~12:15和13:45~14:15采用RF模型,运用全波段的光谱反射率估算大豆叶片SPAD值和LAI可得到较准确的结果。

病毒病侵害下番茄叶片SPAD值高光谱估算模型

以番茄品种‘凯萨’(Kaisa)为试材,于2022年6-8月在南京信息工程大学人工气候室进行盆栽试验。将粉虱作为介体,通过刺吸方式对番茄植株进行病毒传播。设置人工气候箱最高气温/最低气温为28℃/18℃、空气相对湿度为70%(±5个百分点)。利用SOC710光谱仪和SPAD-502叶绿素仪测定番茄苗期叶片不同侵染程度的光谱反射率和叶绿素相对含量(SPAD值),对原始光谱进行一阶导数变换,分析不同侵染程度番茄叶片各高光谱参数和叶绿素相对含量的关系,进而筛选出番茄叶绿素相对含量的敏感波段与特征参数,以期构建叶绿素相对含量的最佳估算模型。结果表明:不同侵染程度下番茄叶片的红边位置不同。在710nm处光谱曲线波峰位置发生左移现象,健康、轻度、中度和重度侵染的番茄叶片最大峰波长分别出现在711nm、700nm、695nm和690nm。通过光谱参数与SPAD值的相关性分析,选出最大极显著相关(P<0.01)的要素R_(531)、R_(816)(正)、R’_(500)和R’_(721)(正)为光谱特征参数。利用光谱特征参数构建光谱指数,与SPAD值呈极显著相关的要素为Dy、λb、λr、Rg、RVI、DVI、PRI、GNDVI、NDVI、m NDI和SAVI。通过逐步回归方法筛选出m NDI、GNDVI、R816、Rg和R’500构建病毒病侵害下番茄叶片SPAD值高光谱估算模型,验证模型的R^(2)为0.835。研究显示,利用光谱技术对病毒病侵害下番茄叶片SPAD值检测是可行的,可为植物病害的早期无损诊断提供参考。

基于叶片SPAD值的华南双季早稻氮肥推荐模型的研究

研究于2021—2022年以合美占和粤晶丝苗2号为供试水稻品种进行了不同氮肥水平(0、60、120、180、240 kg/hm2)的田间试验,分别在水稻分蘖期、拔节期、孕穗期、乳熟期建立基于叶片SPAD值的双季早稻氮肥推荐模型,并对模型推荐的施肥量进行试验验证。结果表明:随着生育期推移,早稻叶片SPAD值呈下降的趋势;同一生育时期,早稻叶片SPAD值随着氮肥用量的增加呈递增的趋势;各生育期水稻产量与叶片SPAD值、施氮量具有显著相关性;以最高产量的90%~95%作为临界值,推知水稻分蘖期、拔节期、孕穗期、乳熟期的临界SPAD值分别为41.2、40.1、33.6、23.5;根据临界SPAD值,计算出基肥、分蘖期、拔节期、孕穗期、乳熟期的氮肥推荐施用量分别为90、27.2、12.7、21.5、6.4 kg/hm2,比常规施氮量减少了37.2 kg/hm2,节省了19%的氮肥,产量比常规施肥稍有增加,但差异未达显著水平。该研究建立的基于叶片SPAD值的早稻氮肥推荐模型可以为水稻全生育期氮素诊断提供参考。