黄河流域棉花品种主要性状时间序列的变化趋势分析

Studies on the Change Trend of Time Series of Main Characters for Cotton Varieties in the Yellow River Valley Region

以棉花黄河区试参试品种的主要性状为原始资料,运用线性回归、谐波及交叉谱分析的原理和技术,对各性状时间序列的线性趋势、周期性变化及相关关系进行了研究。线性回归结果显示,铃重、衣分及霜前花率呈增加趋势,生育期及枯萎病指呈下降趋势。谐波及交叉谱分析结果表明,各序列均存在明显的周期性变化,即皮棉产量、铃重、衣分、生育期及霜前花率以14年为主周期,单株铃数以5.6年为主周期,枯、黄萎病指以28年为主周期;且单株铃数、衣分及霜前花率与皮棉产量在2～3年周期上密切相关,铃重及枯、黄萎病指与皮棉产量在7年周期上密切相关,生育期与皮棉产量在9.3年周期上密切相关。这些结果为育种者掌握育种动态、修正育种目标、提高育种水平及棉花品种主要性状的长期预测提供了科学依据。

Besed on main characters of the cotton varieties that joined the Yellow River Valley Regional Variety Test, according to the technique of linear regression, harmonic analysis and cross spectrum analysis, the linear trend, periodic change and correlated relation of time series of each character for cotton varieties were studied. The results of linear regression showed that there were increasing trend of boll weight, lint percentage and the percentage of seed cotton harvested before frost, the decreasing trend occurred for growing period and Fusarium wilt. However, there are not linear relation between lint yield, boll number per plant, Verticillium wilt and time, respectively. The results of harmonic analysis and cross spectrum analysis showed that there were significantly periodic change of time series for each character, and take 14 years as the the main period for lint yield, boll weight, lint percentage, growing period and the percentage of seed cotton harvested before frost, take 5.6 years as the main period for boll number per plant, take 28 years as the main period for Fusarium and Verticillium wilt; and there were best correlativity of bolls, lint percentage, the percentage of seed cotton harvested before frost with lint yield in 2～3 years, respectively, there were best correlativity of boll weight, Fusarium wilt, Verticillium wilt with lint yield in 7 year period, respectively, there were best correlativity of growing period with lint yield in 9.3 years. These results provided bases for reversion of beeding objective, improvement of beeding standard and long term forecasing of main characters of cotton varieties.