Trends and temperature response in the phenology of crops in Germany

Nicole Estrella, Tim H. Sparks, Annette Menzel

Global Change Biology · 2007 · 285 citations · 31 references

Concepts

TL;DR

The study examines phenology of 78 agricultural and horticultural events in Germany from 1951 to 2004. Researchers detrended the event dates by linear regression on year to remove technological effects, then correlated and regressed the detrended and raw dates with mean monthly and seasonal temperatures. Phenology has advanced on average 1.1–1.3 days per decade, with most events occurring earlier; temperature correlations are largely negative, with spring temperatures driving the response (detrended estimate –3.73 days °C⁻¹ versus full –4.31 days °C⁻¹), and perennial crops showing a higher temperature sensitivity than annual crops.

Abstract

Abstract The phenology of 78 agricultural and horticultural events from a national survey in Germany spanning the years 1951–2004 is examined. The majority of events are significantly earlier now than 53 years ago, with a mean advance of 1.1–1.3 days per decade. The mean trends for ‘true phases’, such as emergence and flowering, of annual and perennial crops are not significantly different, although more trends (78% vs. 46%) are significant for annual crops. We attempt to remove the influence of technological advance or altered farming practices on phenology by detrending the respective time series by linear regression of date (day number) on year. Subsequently, we estimate responses to mean monthly and seasonal temperature by correlation and regression in two ways; with and without removing the year trend first. Nearly all (97%) correlation coefficients are negative, suggesting earlier events in warmer years. Between 82% and 94% of the coefficients with seasonal spring and summer temperatures are significant. The conservative estimate (detrended) of mean temperature response against mean March–May temperature (−3.73 days °C −1 ) is significantly less than the full estimate (−4.31 days °C −1 ), the ‘true’ size of phenological temperature response may lie in between. Perennial crops exhibited a significantly higher temperature response to mean spring temperature than the annual crops.

References

31