【摘要】 本文对黑龙江省67个县(市)近十年的大豆生育季节(5～9月份)内诸月各旬降水、旬积温(≥10℃)与大豆产量的相关关系进行了多元回归分析。分析结果认为全生育季节的持续性高温和季节性少雨是构成黑龙江省大豆生育季节干旱的主要气象因子。而生育季节少雨干旱乃是左右黑龙江省大豆产量的关键。 文章通过全生育季节各旬降水、旬积温(≥10℃)对大豆产量的偏相关、复相关及其决定系数以及多元回归方程和偏回归系数的显著性测定指出:黑龙江省大豆生育季节内主要有四个时期发生干旱与产量的相关密切,即:五月下旬的芽旱,六月中下旬的植株营养生长干旱;七月中下旬的花旱;以及八月中下旬的荚旱。因此,这四个时期可为黑龙江省大豆生育期内对水份的敏感时期。 因此,为最大限度地在黑龙江省增加大豆产量,应注意旱作栽培技术,并应将大豆品种的抗旱性作为黑龙江省大豆育种的主要目标之一。更多还原

【Abstract】 Based upon the precipitation and effective temperature （≥10℃） inthe three dacades of a month in the growing season （from May to Septem-ber）, soybean yield per unit area （jin/per mu） for ten years from 67counties in Heilongjiang Province was analysed by multiple regression.Researchresults indicate that all the multiple regression equations are nearly at thesignificant level （p<0.05） or high significant level（p<0.01）. Partial reg-ression coefficients （b1） of all the multiple regression equations are positive,and the significant rate （at the level of 5% and 1%） is 69.2 percent （exceptblE .July; blE. August）; all the b2 are nearly negative and the signi-ficant rate （at the level of 5 %and1 %） is 64.2 percent. It is knownthat the higher the air temperature is; the lower the precipitation is duringthe total growing season. These are the meteorological factors of droughtin the growing season in Heilongjiang Province, among them the precipitationis the main constraint to influence soybean yield.Partial correlation demonstrates that the partial correlation coefficients(r_yx1·x2) are nearly positive with high significant rate （at the level of 1%）.They are as follows: partial correlation coefficients of the first and thirddecades of May are r_yx1·x2= 0. 3347^** and r_yx1·x2=0.4069^**respectively; oneof the second decade of July is r_yx1·x2=0.5234^**; one of the second decadeof August is ry×1.×2=0.3341^**;one of the third decade of Septemberis r_yx1·x2=0.3489^**.Partial correlation coefficients (r_yx2·x1) are negative, and partialcorrelation coefficients of the second decade of June are r_yx1·x2=-0.04714^**,one of the third decade of June is r_yx2·x1=-0.5082^**; one of the first andsecond decades of July is y_yx2·x1=-0.3423^*and r_yx2·x1=-0.4166^**respectively.It is a high significance （at the level of 1%）. Partial cerrelation coefficientof the first decade in August is r_yx2·x1-0.2363^*, it is a significance （at the level of 5％）. Multiple correlation coefficent(r_yx1·x2)indicated that the precipitationand effective temperature in each decade of a month（from May to September）in the growing season of soybean VS.on its yield in Heilongjiang Provincewas of high significance（at thelevel of 1％）or of general significance（at the level of 5％）.These clearly showed that effects of interactionamong precipitation and effective temperature of each decade of a monthon soybean yield were very obvious. These results suggested to the fullest extent that there be four periods ofdroughts during the growing season of soybean in Heilongjiang Province:e. g. the bud drought of soybean in the third decade of May; seedlingdrought of soybean in the second and third decades of June; flower droughtof soybean in the second and third decades of July and pod drought of soybeanin the second and third decade of August. And therefore,the four periodsare the critical periods for soybean in need of water in its growing season inHeilongjiang Province. In order to Increase soybean yield under the conditions of dronght inits growing season,the main objectives for soybean research in HeilongjiangProvince are to breed the new drought—resistant varieties of soybean and toselect drought resistant cultivars of soybean.更多还原