刘少荣 胡勇 沈茜 马涛 曾厅余 孙东汉

摘 要：為有效防御花期低温冷害，提高苹果产量，更好地服务地方高原特色农业和经济社会发展，选择西南冷凉高地苹果种植区的云南昭通地区作为研究对象，开展苹果始花期预报方法研究。选用2010-2020年苹果始花期资料和同时期气象资料，对花期前的气温、积温、日照等气象因子进行相关性分析，利用逐步回归分析方法，建立单一气温、积温和多因子苹果始花期预报模型共6个，并对始花期开展预测和检验。结果表明：月平均气温和旬平均气温因子与始花期有很好的相关性，通过95%的显著性检验，且平均气温越靠近真实花期，其相关性越显著。通过计算并找出最小积温变异系数，可以确定下限温度（或称生物学零度）和积温时段。运用回代和预测两种方法进行检验表明，建立气温和积温的多因子预报模型，预报误差均小于3 d，而建立气温或者积温的单一因子预报模型，预报误差均大于3 d，多因子模型预报的苹果始花期更靠近真实花期。多因子预报模型提前预测时间可达1周左右，在预防苹果花期低温冷害和提高苹果质量方面起到积极作用。

关键词：苹果;始花期;积温;变异系数

中图分类号：S165

文献标识码：A

文章编号：1008-0457（2021）04-0069-08

国际DOI编码：10.15958/j.cnki.sdnyswxb.2021.04.010

Abstract：In order to effectively prevent low temperature and chilling injury during flowering period，improve apple yield，and better serve local agriculture and economic and social development with plateau characteristic，the Zhaotong area of Yunnan Province，an apple-growing area in the cold highlands of southwestern China，was selected as the research object to carry out the research on the forecasting method of the apple′s first flowering period.Based on the data of early flowering period of apple and the meteorological data of the same period from 2010 to 2020，the correlation analysis of meteorological factors such as temperature，accumulated temperature and sunshine，etc.before the flowering period was carried out.The stepwise regression analysis method was used to establish 6 early flowering period forecast models for apple with a single temperature，accumulated temperature and multi-factor.And this paper also carried out the prediction and inspection of early flowering period.The results showed that monthly average temperature and ten-day average temperature factors had a good correlation with the initial flowering period.The closer the average temperature was to the true flowering period，the more significant the correlation would be.By calculating and finding out the minimum coefficient of variation of accumulated temperature，the lower limit temperature （or biological zero） and accumulated temperature period could be determined.The back-generation and prediction methods showed that the establishment of a multi-factor forecast model for temperature and accumulated temperature had forecast errors less than 3 days，while the establishment of a single-factor forecast model for air temperature or accumulated temperature had forecast errors more than 3 days.Therefore，the multi-factor prediction model was closer to the real flowering period.The prediction time of multi-factor prediction model can predict the time up to about one week in advance，which plays a positive role in the prevention of low temperature and chilling injury and improvement of apple quality.

Keywords：apple;initial flowering period;accumulated temperature;coefficient of variation

西南冷凉高地作为我国苹果四大主产区之一[1]，苹果种植主要分布在云、贵、川高原地区，云贵高原北部的昭通苹果主产地区海拔在1900～2000 m，气候温和，冬无严寒，夏无酷暑[2]，雨量适中，光、热……