【作者】 张丽；

【导师】 董树亭； 王空军；

【作者基本信息】 山东农业大学 ， 作物学， 2008， 博士

【摘要】 本试验于2004年~2007年在山东农业大学玉米科技园和作物生物学国家重点实验室进行。本研究采用田间试验和室内生理生化分析测定相结合的方法。2004年通过品比试验了解了不同品种玉米籽粒容重与产量和品质的关系,结合氮肥试验,以普通型玉米、硬粒型玉米和爆裂型玉米为材料,研究了氮肥对玉米籽粒容重的影响,在此基础上选用农大108(ND108)和费玉4号(FY4)两个有代表性的玉米品种,2005年设置不同的种植密度处理,研究了籽粒容重对种植密度的响应。在容重与产量和品质的相关分析的基础上,以淀粉分析为重点,2006年选用高淀粉玉米费玉3号(FY3)和郑单18(ZD18),以普通型玉米ND108和郑单958(ZD958)为对照,研究了播期对玉米籽粒容重的影响。2007年为补充试验,品种与2006年相同,同一播期设置不同的种植密度。在试验的过程中同时探讨了容重的形成过程及其与籽粒发育的关系;对淀粉着重分析,探讨了容重与淀粉的积累、淀粉粒的形态、大小及排列以及胚乳细胞淀粉粒的粒度分布间的关系,最后总结了不同栽培措施(氮肥、种植密度及播期)对容重的影响。主要研究结果如下:1玉米容重与产量和品质的相关分析探讨了玉米籽粒容重与产量和品质的关系,了解不同品种籽粒容重的差异以及差异的主要来源。1.1籽粒容重与产量和品质的关系容重与产量及相关性状间的相关分析表明,不同类型玉米籽粒容重与千粒重、籽粒比重和产量均呈极显著的正相关,与籽粒漂浮率呈显著负相关。玉米上部、中部和下部的籽粒容重与收获后籽粒水分百分含量均呈显著负相关,随着水分含量的下降,容重呈上升趋势,水分含量每下降一个百分点,容重平均上升4.86、5.33和5.50 g/L。容重与籽粒主要营养成分含量间的相关分析表明,容重与蛋白质含量(r=0.573,P<0.05)和总淀粉含量(r=0.719,P<0.01)呈正相关,与粗脂肪含量和赖氨酸含量呈负相关。通径分析结果表明总淀粉含量对容重的决策系数最大为68.52%,故要提高容重,应该首先着眼于籽粒总淀粉含量的提高。1.2籽粒容重差异的主要来源2004年试验通过聚类分析将山东省最新审定的29个玉米品种划分为三类,高容重型(H-TW)、中容重型(M-TW)和低容重型(L-TW)。玉米容重的差异来源于基因型和试验地区,二者均达极显著水平。受基因型控制的籽粒形状是决定容重的重要因素,不同粒型之间容重差异显著,容重值排序是爆裂型>硬粒型>马齿型。2006年试验结果也表明基因型是决定籽粒容重的重要因素。同时播期、播期×品种的交互作用对籽粒容重及相关性状的影响也达到显著水平。不同粒位之间籽粒容重的差异显著,容重值下部籽粒>上部籽粒>中部籽粒。播期×品种×粒位的相互作用对容重的极显著影响主要归咎于品种因素,因为品种的均方值要远远大于播期和粒位。2.玉米容重与籽粒发育的关系由于容重与籽粒比重间的高度正相关性,用比重在籽粒发育过程的变化趋势来间接反映籽粒灌浆过程中容重的变化是完全可行的。籽粒鲜比重授粉后随着籽粒的发育呈上升趋势,到成熟期趋势稳定,二次曲线模拟方程为y=-0.0007x2+0.0076x+0.9987 (F=77.892,P<0.01,R=0.941)。而干比重在授粉后呈双峰曲线变化。灌浆前期比重处于下降趋势,授粉后16～28d是比重增长的快速时期,灌浆后期稍有下降,到成熟期基本趋于稳定。用三次曲线对籽粒比重授粉后的变化动态进行拟合,方程为y=1.368-0.0232x+0.000696x2-0.000006x3(F=8.172,P<0.01,R=0.671)。粒重、体积和单位体积干物质积累量授粉后的变化趋势均可用Logistic方程较好的模拟,水分百分含量随着籽粒的发育迅速下降,而干鲜体积的差值即籽粒含水量呈单峰曲线变化,在授粉后29天左右达到最大,之后开始下降。籽粒灌浆发育过程中各项指标与比重的回归分析可知,籽粒灌浆快增持续期是比重形成的关键时期,此期间影响籽粒的灌浆将显著影响比重的大小。3.玉米容重与淀粉积累的关系3.1籽粒容重与淀粉粒粒度的分布选用高淀粉型玉米(FY3和ZD18)和普通型玉米(ND108和ZD958)为试材,研究了籽粒胚乳中淀粉粒粒度分布情况,淀粉粒的体积、数目和表面积的分布特性,及其与容重和淀粉含量的关系。结果表明,成熟期玉米籽粒的粒径分布范围为0.37~31.5μm,但其上限值并不完全一致,为26.15~31.51μm,以2μm和15μm为界限,可把淀粉粒分别小型、中型和大型三类。玉米淀粉粒的数目表现为单峰分布,峰值出现在0.829μm,其中小淀粉粒组(<2μm)的淀粉粒数目占总数目的95%左右,是淀粉粒的主要组成部分。玉米淀粉粒的体积表现为双峰分布,峰值分别出现在1.45μm和16.40μm左右;表面积分布表现为三峰分布,峰值分别在1.20μm,4.88μm和14.94μm。小淀粉粒组(<2μm)和中淀粉粒组(2~15μm)所占体积分数高淀粉玉米显著高于普通型玉米;而大淀粉粒组(>15μm)所占的数目、体积和表面积百分比高淀粉玉米相对较低。小、中型淀粉粒组的数目和表面积百分比在品种和年际之间有差异。淀粉粒粒度分布不同粒位间的比较结果表明,上部和下部籽粒大淀粉粒组的体积百分比相对较高,而中部籽粒中小淀粉粒组和中淀粉粒组的体积百分比相对较高。容重与淀粉粒的体积分布的相关分析表明,容重与0.8~2μm,2~10μm和>20μm的淀粉粒体积百分比呈正相关,与其他粒度体积分布呈负相关,但检验均未达到显著水平。而0.8~2μm,2~10μm和10~15μm的淀粉粒体积百分比与籽粒淀粉含量分别呈正相关,其中前两者相关系数达到显著水平,而<0.8μm,15~20μm,和>20μm的淀粉粒体积百分比与籽粒淀粉含量分别呈负相关,但相关系数均不显著。由于容重与淀粉含量的高度正相关性,因此我们可以间接地关注0.8~2μm和2~10μm范围内的淀粉粒度分布情况。3.2淀粉积累动态及淀粉合成相关酶活性灌浆前期,高淀粉玉米和普通型玉米总淀粉含量差异不显著;授粉后20天之后,随着籽粒的发育,直链淀粉含量表现为普通型>高淀粉型,但差异并不显著,支链淀粉和总淀粉含量表现为高淀粉型>普通型,此趋势一直维持到灌浆末期,到成熟期差异达到显著水平。高淀粉玉米的支链淀粉和总淀粉积累速率均大于普通型玉米,而直链淀粉积累速率较低。相对于体积和表面积的淀粉粒的平均径授粉后30天左右呈快速增长趋势,之后基本趋于稳定,而相对于数目而言,淀粉粒平均径授粉后呈下降趋势,到14天左右基本趋于稳定。高淀粉玉米相当于体积和表面积的平均径两年的平均值(12.41μm和6.81μm)均低于普通型玉米(12.98μm和6.40μm),而相对于数目的平均径(1.10μm)大于普通型玉米(1.08μm)。<0.8μm的淀粉粒体积授粉后呈先增加后降低的趋势,0.8-2μm和2-10μm组的淀粉粒一直下降趋势,而>10μm的淀粉粒组的体积一直处于上升的阶段,所有的变化到灌浆中后期都趋于稳定。相应各淀粉粒组的表面积和数目相对百分比与体积相对百分比的变化趋势相一致,品种之间也基本相似。高淀粉玉米籽粒腺苷二磷酸葡萄糖焦磷酸化酶(ADPGPPase)活性、尿苷二磷酸葡萄糖焦磷酸化酶(UDPGPPase)活性和可溶性淀粉合成酶(SSS活性),在灌浆前期低于普通型玉米,而到灌浆中后期开始升高且末期仍然保持较高的活性且下降速率显著低于普通型玉米。整个灌浆过程中高淀粉型玉米的束缚态淀粉合成酶(GBSS)活性均高于普通型玉米。3.3籽粒容重与淀粉粒形态、大小及排列淀粉粒的形状多呈球状、椭圆状和多面体状。根据表面形态的不同,把淀粉粒大致分为两类:淀粉粒表面没有物质包裹的称为“裸露型”;淀粉粒表面有大量的粘性物质(基质蛋白等)包裹,表面还有一些小的“内陷”,称为“非裸露型”。电镜观察玉米籽粒胚乳横断面的淀粉粒大小,直径范围在7.981~20.472μm之间,大多集中在中、大淀粉粒组,而缺少了小淀粉粒组,与淀粉粒粒度分布的分析结果有所差异,这可能与电镜所能观察到的精确度有关。胚乳横断面靠近边缘部和中部的淀粉粒的平均直径大小排序有类似的趋势,均表现为硬粒型>普通型>高淀粉型。淀粉粒的排列方式有差异,大致可以分为疏松、较为紧密和十分紧密三种情况。对于不同类型的玉米品种而言,胚乳横断面淀粉粒排列的紧密程度硬粒型>高淀粉型>普通型。而对于同一品种不同粒位而言,高淀粉型玉米上部、中部和下部籽粒的胚乳横断面淀粉粒排列情况基本相似,而普通型玉米淀粉粒的排列情况为上部、下部>中部。说明淀粉粒的大小并不是影响籽粒容重的决定性因素,而淀粉粒的形态尤其是排列方式的差异才是造成容重差异的内在根本原因。4.栽培措施对玉米容重的影响4.1氮肥和种植密度对籽粒容重的影响施氮肥对玉米容重的影响不大,同一品种不同的施氮量处理之间玉米的容重差异不显著。籽粒容重与种植密度间呈一定程度负相关,随着种植密度的增加,籽粒容重呈下降趋势,但种植密度处理之间差异不显著,说明种植密度对籽粒容重的影响不显著。随着种植密度的增加,ND108和FY4玉米籽粒的单株产量、千粒重、穗行数、行粒数、穗长及穗粗均呈下降趋势。种植密度对籽粒体积以及水分百分含量的影响并无明显规律,而灌浆持续期随着种植密度的增加而缩短,且ND108受影响的程度大于FY4。籽粒比重在灌浆初期种植密度处理之间差异达到显著水平(P<0.01),而随着生育进程的推进,处理之间的差异逐渐缩小,到成熟期籽粒比重D1>D2>D3,但差异不显著。种植密度对淀粉支/直比的影响最大,其次是对可溶性糖含量的影响,而对总淀粉含量的影响最小。随着种植密度的增加,淀粉支/直比和蛋白质含量均呈下降趋势,可溶性糖含量呈上升趋势。种植密度的增加促进了球蛋白含量的提高,但总蛋白含量有所降低。4.2播期对籽粒容重的影响播期对高淀粉玉米(FY3和ZD18)容重的影响大于对普通型玉米(ND108和ZD958)籽粒容重的影响,适当晚播有利于提高玉米籽粒容重。播期直接影响玉米的穗分化和灌浆过程。随着播期的推迟,玉米的出苗期和生育期均有所提前,但品种之间存在一定的差异。播期对穗部性状秃顶长的影响最大,其次是穗长和行粒数。对于粒部性状,播期对籽粒体积的影响最大,其次为粒重,而且播期对粒宽以及粒长/粒宽比值的影响也达到了显著水平。播期对灌浆速率的影响大于对灌浆持续的影响,即播期主要是通过影响玉米籽粒的灌浆速率影响灌浆过程。灌浆期气候参数和持续灌浆期、灌浆平均速率的相关分析结果表明灌浆持续期与日平均气温和降水量呈负相关,相关系数分别为-0.451和-0.583(P<0.05),而与日平均气温和日较差呈正相关,但相关系数均未达到显著水平。而灌浆平均速率与气候参数的关系正好与之相反,灌浆平均速率与气候各参数的相关系数均达显著水平(P<0.05)。其中,与日平均气温和降水量呈正相关,相关系数分别为0.689和0.625,而与日照时数和日较差呈负相关,相关系数分别为-0.608和-0.697。更多还原

【Abstract】 Experiments were conducted in the field of Shandong Agricultural University Research Farm (36°10’19"N, 117°9’03"E) and the state key laboratory of crop biology from 2004 to 2007, in Taian, Shandong Province, China. Field experiment method and physiological- biochemical analysis were used. The variety comparative test was carried out in 2004 in order to discuss the relationships among test weight (TW), yield and quality, at the same time, the nitrogen fertilizer experiment was combined. Three different maize types, normal maize, flint maize and pop maize were used to investigate the effect of nitrogen fertilizer on TW of maize. On this basis, ND108 (normal maize, NM) and FY4 (flint maize) were selected and used in 2005. In this year, three different planting densities were set to study the response of planting density to TW. Based on the correlation analysis in 2004, four maize hybrids ND108, ZD958 (NM), FY3 and ZD18 (high starch maize, HSM) were used in 2006. This study was focused on the starch analysis, associated with the effect of sowing date on TW. Supplement experiment was carried out in 2007.In addition, we discussed the relationship between TW and kernel development focusing on starch analysis. The accumulation of starch, the shape, size, arrangement of starch granules and the starch granule size distribution were studied. Finally, we summarized the effect of cultivation measures (nitrogen fertilizer, planting density and sowing date) on TW of maize. The maize results as followed:1 Correlation Analysis on Test Weight with Yield and Quality in Maize1.1 Relationships among TW, yield and quality of maize The correlation analysis indicated that the TW was significantly and positively correlated with kernel weight, kernel specific gravity and yield. The floating percent was negatively correlated with TW (P<0.05). TW values of apical, middle and basal kernels were significantly and negatively correlated with percent water content after harvest. With the percent water content decreasing, TW was increasing. Each decreasing of percent water would resulting in 4.86, 5.33 and 5.50 g/L increasing of TW in apical, middle and basal maize kernels. There was a negative correlation between TW and gross fat contents.TW was correlated positively and significantly with protein (r=0.573, P<0.05) and starch contents (r=0.719, P<0.01) respectively. The pass analysis also indicated that starch content was determinated factor for TW, because of its largest decision coefficient (R(4)2=68.52%).Therefore, it is more important to improve TW with an eye to the increasing of starch contents on account of the less constraint effect of lysine to starch contents.1.2 Sources of variations of test weight in maizeClustering analysis was used in 2004. Twenty-nine summer maize hybrids released for commercial production in Shandong province were divided into three types: high-test weight (H-TW), medium-test weight type (M-TW) and low-test weight type (L-TW). Both sources of variations of genotypes and experiment locations were significant. Kernel type was the first important factor to determine the TW, and the difference of TW between kernel types was significant. The value was ordered as Pop> Flint >Dent. Results of 2006 indicated that seeding date and genotype played important and determinant roles on TW. The mean squares for sowing date-by-hybrid interaction, position factor, and hybrid-by-position interaction were significant different both for TW and kernel size traits. Basal kernels had the highest TW, apical kernels had intermediate, and middle kernels had lowest TW. Finally, a significant sowing date-by-hybrid-by-position interaction was observed for TW even though the meaningfulness of the interaction could be ascribed to hybrid factor, because the mean square was much greater than that of the seeding date and position factor.2. Correlation between Test Weight and Kernel DevelopmentDue to the positively and significantly correlation between test weight and kernel specific gravity, it was reasonable to discuss the changes of specific gravity instead of test weight during grain filling. Fresh and dry specific gravities of maize kernels were studied after pollination till to the maturity. The fresh specific gravity increased regularly and trended to stable in the maturity. Quadratic equation of the simulation curve was y=1.368-0.0232x+0.000696x2-0.000006x3, (F=8.172, P<0.01, R=0.6706). However, the dry specific gravity shoed two-peak curve. Cubic equation of the simulation curve was y=1.368-0.0232x+0.000696x2-0.000006x3(F=8.172, P<0.01, R=0.6706)。100-kernel weight, kernel volume and dry matter accumulation per volume changing curves could be simulated by Logistic equations suitably. The percent water content decreased rapidly with the development of kernels. The⊿volume per kernel showed typical single curve, reached its maximum at about 29 days after pollination, and decreased later. Regression analysis between specific gravity and each index of grain filling indicated that the fast-increasing period of filling was the key period of specific gravity forming. During the period, measures which influenced grain filling greatly would affect specific gravity similarly.3. Correlation between Test Weight and Starch Accumulation3.1 Test weight and starch granule size distributionFour maize hybrids ND108, ZD958 (NM), FY3 and ZD18 (HSM) were used in 2006. The purpose of this study was focused on starch granule size distribution and the relationship between starch granule size and grain quality properties. The starch granule in matured grain was 0.37-31.5μm in diameter. Taking 2μm and 15μm as limit, we divided the starch particles into three types: small starch granule group (SSG, <2μm), middle starch granule group (MSG, 2μm-15μm) and large starch granule group (LSG, >15μm). The distribution showed typical single peak curve in number of starch granule, with the peak value was about 0.829μm. The SSG granule accounted for over 95% of the total starch granule. The distribution showed two-peak curve in starch granule volume, and the peak value occurred at 1.45μm and 16.4μm around. The distribution showed typical three-peak curve in starch granule surface area, and the peak value occurred at 1.2μm, 4.88μm and 14.94μm around.The percent of volume of SSG and MSG were higher than that of LSG in the high starch maize, but in the normal maize, the percent of volume, number and surface area of LSG were higher than those in high starch maize. The percent of number and surface area of SSG and MSG were different in hybrids and years. In addition, the percent of volume of LSG was higher than that of SSG and MSG in the apical and basal maize kernels, but on the contrary in middle maize kernels.Correlation analysis indicated that test weight was positively correlated with the volume of 0.8-2μm, 2-10μm and >20μm starch granules, respectively, but negatively correlated to other size ranges, and all the correlation coefficients were not significant. The starch content was positively correlated with the volume of 0.8-2μm(r=0.777, P<0.05), 2-10μm (r=0.735, P<0.05) and >20μm starch granules, but on the contrary to the protein content. The content of starch and protein had no correlation with volume percentage of starch granules with other size ranges. Therefore, based on the positively and significantly correlation between TW and starch content, we would focus on the 0.8-2μm and 2-10μm starch granules.3.2 Starch accumulation and enzyme activities responsible for starch biosynthesis in developing grainsAt early grain filling stage, there was no difference between high starch maize and normal maize in percent starch content. After about 20 DAP, with the increasing of kernel weight, the content of amylose of high starch maize was higher than normal maize, but not significantly. At maturity, the contents of amylopectin and starch were higher than those of normal maize significantly. High starch maize had higher accumulation rates of amylopection and starch and lower accumulation rate of amylase than those of normal maize. The average diameters of starch granules relative to volume and surface area increased quickly during 30 DAP then increased slowly at last stage of growth. Comparatively, the average diameter of starch granules relative to number decreased rapidly in the early and stabilized at about 14 days after pollination. High starch maize had lower average diameter (12.41μm and 6.81μm) and larger average diameter (1.10μm) of starch granules, which were relative to volume, surface area and number respectively than those of normal maize (12.98μm, 6.40μm and 1.08μm). The volume percent of starch granules of each size range changed differently in the grain filling. The volume of <0.8μm starch granules increased in the early then decreased. The volume of 0.8-2μm and 2-10μm starch granules decreased in the early, and the volume of >10μm starch granules increased in the early, all the changes stabilized in the middle and later period of grain filling. The number and surface area of starch granules of each size range had similar trends. The activities of adenosine diphosphate glucose pyrophosphorylase (ADPGPPase), uridine diphosphate glucose pyrophosphorylase (UDPGPPase), and soluble starch synthase (SSS) of high starch maize were lower than those of normal maize, increased in the middle and later grain filling stage, then kept relative and high activities at last stage of growth. Comparatively, the activity of granule-bound starch synthase (GBSS) of high starch maize was lower than that of normal maize during the whole grain filling stage.3.3 Test weight and the shape, size and arrangement of starch granulesThe shape of starch granules was spherical, elliptical or polyhedral. According to the surface traits, we divided the starch granules into two types: the“uncovered endosperm starch”with smooth surface and the“covered endosperm starch”attached with much matrix protein and matrix granules. The average starch granule diameter observed with Scanning Electron Microscope (SEM) was within the range from 7.891μm to 20.472μm. That is to say, the middle and large starch granule groups were in the majority, and it was lack of the small starch granule group. This result was different from the analysis with starch granule size distribution. It was possibly resulted from the different accuracy of SEM relative to particle distribution analysis. Flint type maize had the highest average starch granule diameter either in the central or in the edge endosperm, normal maize had intermediate, and high starch maize had the lowest average starch granule diameter.Starch granule arrangement had three types: loosely, tightly relatively and tightly. As to different maize hybrids, starch granules of flint type maize in endosperm arranged tightly, the next was high starch maize, and starch granules of normal maize arranged loosely. In addition, as to the different kernel positions with the same hybrid, the apical, middle and basal kernel of high starch maize had the similar arrangement of starch granules. Comparatively, as to normal maize, starch granules arrangement of apical, middle and basal kernels in endosperm had a little difference. Thus, the size of starch granules was not the decisive factor to the difference of test weight, but the shape and structure of starch granules, especially the varies of arrangement were the internal basic reasons.4 The Effect of Cultivation Measures on Test Weight in Maize4.1 The effect of nitrogen fertilizer and planting density on test weight The effect of nitrogen fertilizer on test weight in maize was not significant. Test weight was negatively correlated with planting density to some extent. Test weight decreased with planting density increased. But there was no significant difference among treatments. Thus, we can conclude that the effect of planting density on test weight in maize was also not significant.With the increasing of planting density, the yield per plant and its component factors of ND108 and FY4 all decreased. Only the ear diameter and row number kept relatively stable. Grain filling was affected by planting density, with the grain filling duration decreased. At the early grain filling stage, specific gravity was significantly different form each treatment. With the kernel weight increasing, the differences declined gradually. At the maturity, the specific gravity value was D1>D2>D3, but no significant difference.The ratio of amylase and amylopectin was sensitivest to planting density, the soluble sugar content was intermediate, and planting density had the smallest influence at the total starch content. As planting density increasing, the ratio of amylase and amylopectin and crude protein content decreased, but soluble sugar content increased. Increasing planting density was beneficial to the promotion of globulin content, but the total protein content decreased.4.2 The effect of sowing date on test weightSowing date had more impact in high starch maize (FY3 and ZD18) than on normal maize (ND108 and ZD958). Proper and late sowing was beneficial to the promotion of test weight.Sowing date affected spike differentiation and grain filling of maize directly. With delaying of sowing date, seeding and growth periods were shorter, but different in hybrids. The bare tip length was sensitivest to sowing date, the next were ear length and kernel numbers per row. As to kernel features, kernel volume was affected greater by sowing date than kernel weight, and the effect of sowing date on the ratio of kernel length and kernel width was significant.Sowing date influenced grain filling mainly through affecting the grain filling rate rather than grain filling duration. Correlation analysis indicated that grain filling duration was significantly and negatively related to daily mean temperature and precipitation. The correlation coefficients were -0.451 and -0.583, respectively. On the contrary, there were significantly and positively correlation between the average filling rate with daily mean temperature (r=0.689) and precipitation (r=0.625), however, the sunshine hours (r=-0.608, P<0.05) and daily range (r=-0.697, P<0.05) were negatively correlated with the average filling rate.更多还原